Road/Lane Detection Evaluation 2013


This benchmark has been created in collaboration with Jannik Fritsch and Tobias Kuehnl from Honda Research Institute Europe GmbH. The road and lane estimation benchmark consists of 289 training and 290 test images. It contains three different categories of road scenes:

  • uu - urban unmarked (98/100)
  • um - urban marked (95/96)
  • umm - urban multiple marked lanes (96/94)
  • urban - combination of the three above

Ground truth has been generated by manual annotation of the images and is available for two different road terrain types: road - the road area, i.e, the composition of all lanes, and lane - the ego-lane, i.e., the lane the vehicle is currently driving on (only available for category "um"). Ground truth is provided for training images only.

We evaluate road and lane estimation performance in the bird's-eye-view space. For the classical pixel-based evaluation we use established measures as discussed in our ITSC 2013 publication. MaxF: Maximum F1-measure, AP: Average precision as used in PASCAL VOC challenges, PRE: Precision, REC: Recall, FPR: False Positive Rate, FNR: False Negative Rate (the four latter measures are evaluated at the working point MaxF), F1: F1 score, HR: Hit rate. For the novel behavior-based evaluation a corridor with the vehicle width (2.2m) is fitted to the lane estimation processing result and evaluation is performed for 3 different distance values: 20 m, 30 m, and 40 m. We refer to our ITSC 2013 publication for more details.
IMPORTANT NOTE: On 09.02.2015 we have improved the accuracy of the ground truth and re-calculated the results for all methods. Please download the devkit and the dataset with the improved ground truth for training again, if you have downloaded the files prior to 09.02.2015. Please consider reporting these new number for all future submissions. The last leaderboards right before the changes can be found here!

Additional information used by the methods
  • Stereo: Method uses left and right (stereo) images
  • Laser Points: Method uses point clouds from Velodyne laser scanner
  • GPS: Method uses GPS information
  • Additional training data: Use of additional data sources for training (see details)

Road Estimation Evaluation

UM_ROAD


Method Setting Code MaxF AP PRE REC FPR FNR Runtime Environment
1 UNV 96.69 % 92.41 % 97.38 % 96.01 % 1.18 % 3.99 % 1.2 s GPU @ 3.0 Ghz (Python + C/C++)
2 NF2CNN
This method makes use of Velodyne laser scans.
96.09 % 88.40 % 94.11 % 98.16 % 2.80 % 1.84 % .006 s GPU @ 3.5 Ghz (Python)
3 KRSF 96.02 % 93.60 % 95.61 % 96.44 % 2.02 % 3.56 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
4 KRS 95.89 % 93.51 % 95.79 % 95.99 % 1.92 % 4.01 % 0.3 s GPU @ 2.5 Ghz (Python)
5 iDST-VT 95.87 % 93.32 % 96.03 % 95.71 % 1.80 % 4.29 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
6 YhY code 95.80 % 89.11 % 94.89 % 96.73 % 2.38 % 3.27 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
7 LidCamNet
This method makes use of Velodyne laser scans.
95.62 % 93.54 % 95.77 % 95.48 % 1.92 % 4.52 % 0.15 s GPU @ 2.5 Ghz (Python)
8 DFFA 95.58 % 89.30 % 95.10 % 96.06 % 2.25 % 3.94 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
9 MVnet
This method makes use of Velodyne laser scans.
95.45 % 91.49 % 97.51 % 93.49 % 1.09 % 6.51 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
10 RSNet 95.28 % 92.43 % 95.22 % 95.35 % 2.18 % 4.65 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
11 DCCN
This method makes use of Velodyne laser scans.
95.18 % 92.44 % 94.69 % 95.68 % 2.45 % 4.32 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
12 TDCac1 CNN 94.86 % 89.62 % 95.45 % 94.28 % 2.05 % 5.72 % .093 s 1 core @ 1.0 Ghz (C/C++)
13 RSNet- 94.84 % 92.83 % 94.32 % 95.37 % 2.62 % 4.63 % 0.07 s GPU @ 2.5 Ghz (Python)
14 baseline 94.80 % 92.80 % 94.35 % 95.25 % 2.60 % 4.75 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
15 IDA-Fusion
This method makes use of Velodyne laser scans.
94.77 % 88.03 % 93.71 % 95.86 % 2.93 % 4.14 % 0.1 s 4 cores @ 3.5 Ghz (C/C++)
16 RBNet 94.77 % 91.42 % 95.16 % 94.37 % 2.19 % 5.63 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
17 WSLGAN 94.73 % 89.22 % 95.01 % 94.45 % 2.26 % 5.55 % 800ms GPU @ 1.5 Ghz (Python)
18 MMN 94.72 % 92.51 % 94.84 % 94.60 % 2.34 % 5.40 % 0.1 s GPU @ 2.5 Ghz (C/C++)
19 KRS 94.69 % 93.40 % 94.72 % 94.67 % 2.41 % 5.33 % 1 s GPU @ 2.5 Ghz (Python)
20 RSNet2 94.65 % 92.54 % 94.45 % 94.85 % 2.54 % 5.15 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
21 SSLGAN 94.62 % 89.50 % 95.32 % 93.93 % 2.10 % 6.07 % 700ms GPU @ 1.5 Ghz (Python)
22 FNETMS 94.51 % 92.72 % 94.92 % 94.11 % 2.30 % 5.89 % 0.04 s GPU @ 2.0 Ghz (Python + C/C++)
23 DEEP-DIG 94.16 % 93.41 % 95.02 % 93.32 % 2.23 % 6.68 % 0.14 s GPU @ 3.5 Ghz (Python + C/C++)
J. Muñoz-Bulnes, C. Fernandez, I. Parra, D. Fernández-Llorca and M. Sotelo: Deep Fully Convolutional Networks with Random Data Augmentation for Enhanced Generalization in Road Detection. Workshop on Deep Learning for Autonomous Driving on IEEE 20th International Conference on Intelligent Transportation Systems 2017.
24 FusionNet
This method uses stereo information.
94.15 % 92.26 % 95.18 % 93.14 % 2.15 % 6.86 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
25 StixelNet II 94.05 % 85.85 % 91.30 % 96.98 % 4.21 % 3.02 % 1.2 s 1 core @ 3.0 Ghz (Matlab + C/C++)
N. Garnett, S. Silberstein, S. Oron, E. Fetaya, U. Verner, A. Ayash, V. Goldner, R. Cohen, K. Horn and D. Levi: Real-time category-based and general obstacle detection for autonomous driving. 5th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving (CVRSUAD'17, IEEE-ICCV 2017 Workshop) 2017.
26 MultiNet code 93.99 % 93.24 % 94.51 % 93.48 % 2.47 % 6.52 % 0.17 s GPU @ 2.5 Ghz (Python + C/C++)
M. Teichmann, M. Weber, J. Zoellner, R. Cipolla and R. Urtasun: MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving. CoRR 2016.
27 FDN 93.99 % 93.29 % 94.53 % 93.46 % 2.47 % 6.54 % 0.2 s GPU 1 core @ 2.5 Ghz (Python)
28 CoDNN 93.90 % 92.86 % 94.51 % 93.29 % 2.47 % 6.71 % 0.01 s 1 core @ 2.5 Ghz (Python + C/C++)
29 FuseNet
This method uses stereo information.
93.86 % 93.34 % 94.49 % 93.23 % 2.48 % 6.77 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
30 FCN-GCBs 93.86 % 86.62 % 92.15 % 95.62 % 3.71 % 4.38 % 0.08 s GPU @ 2.5 Ghz (C/C++)
31 ResNetPK
This method makes use of Velodyne laser scans.
93.78 % 89.01 % 94.78 % 92.81 % 2.33 % 7.19 % 0.4s GPU @ 1.5 Ghz (Python)
32 MBN 93.77 % 85.59 % 91.02 % 96.69 % 4.34 % 3.31 % 0.16 s GPU @ 2.5 Ghz (Python)
33 DDN 93.65 % 88.55 % 94.28 % 93.03 % 2.57 % 6.97 % 2 s GPU @ 2.5 Ghz (Python + C/C++)
R. Mohan: Deep Deconvolutional Networks for Scene Parsing. 2014.
34 wt 93.27 % 92.92 % 93.20 % 93.34 % 3.10 % 6.66 % 0.1 s GPU @ 1.0 Ghz (Python)
35 RDSN 92.77 % 87.54 % 93.16 % 92.39 % 3.09 % 7.61 % 0.25 s GPU @ 2.5 Ghz (Python)
36 LoDNN
This method makes use of Velodyne laser scans.
92.75 % 89.98 % 90.09 % 95.58 % 4.79 % 4.42 % 18 ms GPU @ 2.5 Ghz (Torch)
L. Caltagirone, S. Scheidegger, L. Svensson and M. Wahde: Fast LIDAR-based Road Detection Using Fully Convolutional Neural Networks. IEEE Intelligent Vehicles Symposium 2017.
37 SUNet 92.63 % 86.51 % 92.03 % 93.24 % 3.68 % 6.76 % 0.018s
38 VGGFCN-6D
This method makes use of Velodyne laser scans.
92.33 % 87.34 % 92.95 % 91.71 % 3.17 % 8.29 % .006 s GPU @ 3.5 Ghz (Python)
39 DFN 92.32 % 88.26 % 91.79 % 92.85 % 3.78 % 7.15 % 0.25 s GPU @ >3.5 Ghz (Python)
40 RSNetVGG 92.26 % 92.51 % 93.92 % 90.65 % 2.67 % 9.35 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
41 Up-Conv-Poly code 92.20 % 88.85 % 92.57 % 91.83 % 3.36 % 8.17 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
42 HID-LS
This method makes use of Velodyne laser scans.
92.03 % 83.73 % 88.97 % 95.31 % 5.38 % 4.69 % 0.5 s 4 cores @ 3.0 Ghz (C/C++)
43 MixedCRF
This method makes use of Velodyne laser scans.
91.57 % 84.68 % 90.02 % 93.19 % 4.71 % 6.81 % 6s 1 core @ 2.5 Ghz (Matlab + C/C++)
44 LiDAR-SPHnet
This method makes use of Velodyne laser scans.
91.50 % 81.98 % 87.05 % 96.42 % 6.53 % 3.58 % 0.14 s GPU @ 1.5 Ghz (Matlab)
45 FTP 91.20 % 90.60 % 91.11 % 91.29 % 4.06 % 8.71 % 0.28 s GPU @ 2.5 Ghz (C/C++)
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
46 HybridCRF
This method makes use of Velodyne laser scans.
90.99 % 85.26 % 90.65 % 91.33 % 4.29 % 8.67 % 1.5 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, R. Wang, B. Dai, Y. Fang, D. Liu and T. Wu: Hybrid Conditional Random Field based Camera-LIDAR Fusion for Road Detection. Information Sciences. in press.
47 NNP
This method uses stereo information.
90.50 % 87.95 % 91.43 % 89.59 % 3.83 % 10.41 % 5 s 4 cores @ 2.5 Ghz (Matlab)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
48 Up-Conv 90.48 % 88.20 % 91.30 % 89.67 % 3.90 % 10.33 % 0.05 s GPU @ 2.5 Ghz (C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
49 TFSeg 90.12 % 89.85 % 88.00 % 92.36 % 5.74 % 7.64 % 0.07 s GPU @ 1.0 Ghz (Python)
50 HIM 90.07 % 79.98 % 90.79 % 89.35 % 4.13 % 10.65 % 7 s >8 cores @ 2.5 Ghz (Python + C/C++)
D. Munoz, J. Bagnell and M. Hebert: Stacked Hierarchical Labeling. European Conference on Computer Vision (ECCV) 2010.
51 LidarHisto
This method makes use of Velodyne laser scans.
code 89.87 % 83.03 % 91.28 % 88.49 % 3.85 % 11.51 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Chen, J. Yang and H. Kong: Lidar-histogram for fast road and obstacle detection. 2017 IEEE International Conference on Robotics and Automation (ICRA) 2017.
52 FusedCRF
This method makes use of Velodyne laser scans.
89.55 % 80.00 % 84.87 % 94.78 % 7.70 % 5.22 % 2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, T. Hu and T. Wu: CRF based Road Detection with Multi-Sensor Fusion. Intelligent Vehicles Symposium (IV) 2015.
53 LWDS 89.50 % 86.11 % 90.59 % 88.44 % 4.19 % 11.56 % 0.07 s GPU @ 2.5 Ghz (Python)
54 FCN-LC 89.36 % 78.80 % 89.35 % 89.37 % 4.85 % 10.63 % 0.03 s GPU Titan X
C. Mendes, V. Frémont and D. Wolf: Exploiting Fully Convolutional Neural Networks for Fast Road Detection. IEEE Conference on Robotics and Automation (ICRA) 2016.
55 CB 88.89 % 82.17 % 87.26 % 90.58 % 6.03 % 9.42 % 2 s 1 core @ 3.4 Ghz (Python) + GPU
C. Mendes, V. Frémont and D. Wolf: Vision-Based Road Detection using Contextual Blocks. 2015.
56 VD 88.87 % 82.04 % 93.31 % 84.83 % 2.77 % 15.17 % 4 s 1 core @ 3.0 Ghz (Matlab + C/C++)
57 HFM 88.83 % 80.33 % 85.23 % 92.75 % 7.32 % 7.25 % 5 s 2 cores @ 2.0 Ghz (C/C++)
58 DVFCN 88.64 % 91.37 % 89.20 % 88.10 % 4.86 % 11.90 % 0.07 s GPU @ 2.5 Ghz (Python)
59 ResAXN 88.46 % 91.06 % 90.10 % 86.88 % 4.35 % 13.12 % 0.06 s GPU @ 1.5 Ghz (Python)
60 SPRAY 88.14 % 91.24 % 88.60 % 87.68 % 5.14 % 12.32 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
61 ProbBoost
This method uses stereo information.
87.48 % 80.13 % 85.02 % 90.09 % 7.23 % 9.91 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: A probabilistic distribution approach for the classification of urban roads in complex environments. Workshop on Modelling, Estimation, Perception and Control of All Terrain Mobile Robots on IEEE International Conference on Robotics and Automation (ICRA) 2014.
62 LWD 87.39 % 88.85 % 84.99 % 89.92 % 7.24 % 10.08 % 0.07 s GPU @ 2.5 Ghz (Python)
63 MAP 87.33 % 89.62 % 85.77 % 88.95 % 6.73 % 11.05 % 0.28s GPU
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
64 CN24 86.32 % 89.19 % 87.80 % 84.89 % 5.37 % 15.11 % 30 s >8 cores @ 2.5 Ghz (C/C++)
C. Brust, S. Sickert, M. Simon, E. Rodner and J. Denzler: Convolutional Patch Networks with Spatial Prior for Road Detection and Urban Scene Understanding. VISAPP 2015 - Proceedings of the 10th International Conference on Computer Vision Theory and Applications, Berlin, Germany, 11-14 March, 2015 2015.
65 GRES3D+VELO
This method makes use of Velodyne laser scans.
85.43 % 83.04 % 82.69 % 88.37 % 8.43 % 11.63 % 60 ms 4 cores @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
66 StixelNet 85.33 % 72.14 % 81.21 % 89.89 % 9.48 % 10.11 % 1 s GPU @ 2.5 Ghz (C/C++)
D. Levi, N. Garnett and E. Fetaya: StixelNet: A Deep Convolutional Network for Obstacle Detection and Road Segmentation.. 26TH British Machine Vision Conference (BMVC) 2015.
67 SPlane + BL
This method uses stereo information.
85.23 % 88.66 % 83.43 % 87.12 % 7.89 % 12.88 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
68 geo+gpr+crf
This method uses stereo information.
85.13 % 72.24 % 81.33 % 89.29 % 9.34 % 10.71 % 30 s 1 core @ 2.0 Ghz (C/C++)
Z. Xiao, B. Dai, H. Li, T. Wu, X. Xu, Y. Zeng and T. Chen: Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion. International Journal of Advanced Robotic Systems 2017.
69 Strait 84.28 % 87.89 % 83.66 % 84.91 % 7.56 % 15.09 % 69 ms GPU @ K20
70 RES3D-Velo
This method makes use of Velodyne laser scans.
83.81 % 73.95 % 78.56 % 89.80 % 11.16 % 10.20 % 0.36 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Wolf and C. Stiller: Road Terrain Detection: Avoiding Common Obstacle Detection Assumptions Using Sensor Fusion. Intelligent Vehicles Symposium (IV) 2014.
71 SCRFFPFHGSP
This method uses stereo information.
83.73 % 72.89 % 82.13 % 85.39 % 8.47 % 14.61 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
72 GRES3D+SELAS
This method uses stereo information.
83.69 % 84.61 % 78.31 % 89.88 % 11.35 % 10.12 % 110 ms 4 core @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
73 HistonBoost
This method uses stereo information.
83.68 % 72.79 % 82.01 % 85.42 % 8.54 % 14.58 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: Comprehensive Performance Analysis of Road Detection Algorithms Using the Common Urban Kitti-Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application on IEEE Intelligent Vehicles Symposium (IV) 2014.
74 FRS_SP 83.22 % 72.94 % 77.11 % 90.39 % 12.23 % 9.61 % 0.21 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
75 SegNet 82.17 % 76.46 % 84.03 % 80.40 % 6.97 % 19.60 % 0.01 s 8 cores @ 2.5 Ghz (C/C++)
76 LKW 82.01 % 85.26 % 78.83 % 85.46 % 10.46 % 14.54 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
77 SP-SS 81.60 % 69.62 % 78.13 % 85.40 % 10.89 % 14.60 % 0.01 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
78 PGM-ARS 80.97 % 69.11 % 77.51 % 84.76 % 11.21 % 15.24 % 0.05 s i74700MQ @ 2.1Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: Fast Pixelwise Road Inference Based on Uniformly Reweighted Belief Propagation . Proc. IEEE Intelligent Vehicles Symposium 2015.
79 Pos-ex code 79.47 % 85.99 % 78.78 % 80.17 % 9.84 % 19.83 % 120 ms GPU(K20) @ 0.7 Ghz (Matlab + Caffe)
80 RES3D-Stereo
This method uses stereo information.
78.98 % 80.06 % 75.94 % 82.27 % 11.88 % 17.73 % 0.7 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Gomes and D. Wolf: Road estimation with sparse 3D points from stereo data. Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on 2014.
81 BM
This method uses stereo information.
78.90 % 66.06 % 69.53 % 91.19 % 18.21 % 8.81 % 2 s 2 cores @ 2.5 Ghz (Matlab)
B. WANG, V. Fremont and S. Rodriguez Florez: Color-based Road Detection and its Evaluation on the KITTI Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application, IEEE Intelligent Vehicles Symposium 2014.
82 CNN_LSTM 78.42 % 69.11 % 83.28 % 74.09 % 6.78 % 25.91 % 54 ms GPU @ 2.0 Ghz (Python + C/C++)
83 SPlane
This method uses stereo information.
78.19 % 76.41 % 72.03 % 85.50 % 15.13 % 14.50 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
84 SRF 76.43 % 83.24 % 75.53 % 77.35 % 11.42 % 22.65 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, D. Zhao and T. Wu: Monocular Road Detection Using Structured Random Forest. Int J Adv Robot Syst 2016.
85 CN24 76.28 % 79.29 % 72.44 % 80.55 % 13.96 % 19.45 % 20 s >8 cores @ 2.5 Ghz (C/C++)
C. Brust, S. Sickert, M. Simon, E. Rodner and J. Denzler: Convolutional Patch Networks with Spatial Prior for Road Detection and Urban Scene Understanding. VISAPP 2015 - Proceedings of the 10th International Conference on Computer Vision Theory and Applications, Berlin, Germany, 11-14 March, 2015 2015.
86 CN 73.69 % 76.68 % 69.18 % 78.83 % 16.00 % 21.17 % 2 s 1 core @ 2.5 Ghz (C/C++)
J. Alvarez, T. Gevers, Y. LeCun and A. Lopez: Road Scene Segmentation from a Single Image. ECCV 2012 2012.
87 ARSL-AMI 71.97 % 61.04 % 78.03 % 66.79 % 8.57 % 33.21 % 0.05 s 4 cores @ 2.5 Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: CRF-based semantic labeling in miniaturized road scenes . Proc. IEEE Intelligent Transportation Systems 2014.
88 ANN
This method uses stereo information.
62.83 % 46.77 % 50.21 % 83.91 % 37.91 % 16.09 % 3 s 1 core @ 3.0 Ghz (C/C++)
G. Vitor, D. Lima, A. Victorino and J. Ferreira: A 2D/3D Vision Based Approach Applied to Road Detection in Urban Environments. Intelligent Vehicles Symposium (IV), 2013 IEEE 2013.
89 VAP 59.23 % 42.05 % 44.44 % 88.75 % 50.55 % 11.25 % 1 s 1 core @ 2.5 Ghz (Matlab)
ERROR: Wrong syntax in BIBTEX file.
Table as LaTeX | Only published Methods

UMM_ROAD


Method Setting Code MaxF AP PRE REC FPR FNR Runtime Environment
1 NF2CNN
This method makes use of Velodyne laser scans.
97.77 % 93.31 % 97.41 % 98.13 % 2.87 % 1.87 % .006 s GPU @ 3.5 Ghz (Python)
2 iDST-VT 97.70 % 95.54 % 97.58 % 97.83 % 2.67 % 2.17 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
3 YhY code 97.42 % 93.08 % 97.15 % 97.68 % 3.15 % 2.32 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
4 KRSF 97.34 % 95.58 % 97.42 % 97.26 % 2.83 % 2.74 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
5 UNV 97.34 % 94.23 % 97.52 % 97.16 % 2.71 % 2.84 % 1.2 s GPU @ 3.0 Ghz (Python + C/C++)
6 KRS 97.27 % 95.55 % 97.19 % 97.34 % 3.09 % 2.66 % 0.3 s GPU @ 2.5 Ghz (Python)
7 DFFA 97.26 % 92.75 % 96.79 % 97.74 % 3.56 % 2.26 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
8 DCCN
This method makes use of Velodyne laser scans.
97.10 % 95.31 % 96.64 % 97.56 % 3.73 % 2.44 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
9 LidCamNet
This method makes use of Velodyne laser scans.
97.08 % 95.51 % 97.28 % 96.88 % 2.98 % 3.12 % 0.15 s GPU @ 2.5 Ghz (Python)
10 MVnet
This method makes use of Velodyne laser scans.
96.99 % 93.94 % 98.10 % 95.90 % 2.04 % 4.10 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
11 WSLGAN 96.95 % 92.87 % 96.92 % 96.98 % 3.39 % 3.02 % 800ms GPU @ 1.5 Ghz (Python)
12 baseline 96.88 % 95.39 % 96.36 % 97.40 % 4.04 % 2.60 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
13 RSNet 96.85 % 95.26 % 96.79 % 96.91 % 3.54 % 3.09 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
14 KRS 96.84 % 95.49 % 96.70 % 96.98 % 3.64 % 3.02 % 1 s GPU @ 2.5 Ghz (Python)
15 SSLGAN 96.72 % 92.99 % 97.05 % 96.40 % 3.22 % 3.60 % 700ms GPU @ 1.5 Ghz (Python)
16 RSNet2 96.60 % 95.28 % 96.57 % 96.63 % 3.78 % 3.37 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
17 CoDNN 96.56 % 95.33 % 96.41 % 96.71 % 3.96 % 3.29 % 0.01 s 1 core @ 2.5 Ghz (Python + C/C++)
18 RSNet- 96.40 % 95.34 % 96.59 % 96.22 % 3.74 % 3.78 % 0.07 s GPU @ 2.5 Ghz (Python)
19 IDA-Fusion
This method makes use of Velodyne laser scans.
96.37 % 92.66 % 96.69 % 96.06 % 3.61 % 3.94 % 0.1 s 4 cores @ 3.5 Ghz (C/C++)
20 StixelNet II 96.22 % 91.24 % 95.13 % 97.33 % 5.48 % 2.67 % 1.2 s 1 core @ 3.0 Ghz (Matlab + C/C++)
N. Garnett, S. Silberstein, S. Oron, E. Fetaya, U. Verner, A. Ayash, V. Goldner, R. Cohen, K. Horn and D. Levi: Real-time category-based and general obstacle detection for autonomous driving. 5th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving (CVRSUAD'17, IEEE-ICCV 2017 Workshop) 2017.
21 FDN 96.19 % 95.26 % 95.49 % 96.89 % 5.03 % 3.11 % 0.2 s GPU 1 core @ 2.5 Ghz (Python)
22 MultiNet code 96.15 % 95.36 % 95.79 % 96.51 % 4.67 % 3.49 % 0.17 s GPU @ 2.5 Ghz (Python + C/C++)
M. Teichmann, M. Weber, J. Zoellner, R. Cipolla and R. Urtasun: MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving. CoRR 2016.
23 MMN 96.15 % 94.98 % 96.07 % 96.22 % 4.33 % 3.78 % 0.1 s GPU @ 2.5 Ghz (C/C++)
24 MBN 96.14 % 91.43 % 95.33 % 96.96 % 5.22 % 3.04 % 0.16 s GPU @ 2.5 Ghz (Python)
25 wt 96.11 % 95.44 % 96.08 % 96.14 % 4.31 % 3.86 % 0.1 s GPU @ 1.0 Ghz (Python)
26 TDCac1 CNN 96.11 % 92.62 % 96.65 % 95.57 % 3.64 % 4.43 % .093 s 1 core @ 1.0 Ghz (C/C++)
27 FNETMS 96.08 % 94.95 % 95.56 % 96.60 % 4.93 % 3.40 % 0.04 s GPU @ 2.0 Ghz (Python + C/C++)
28 RBNet 96.06 % 93.49 % 95.80 % 96.31 % 4.64 % 3.69 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
29 LoDNN
This method makes use of Velodyne laser scans.
96.05 % 95.03 % 95.79 % 96.31 % 4.66 % 3.69 % 18 ms GPU @ 2.5 Ghz (Torch)
L. Caltagirone, S. Scheidegger, L. Svensson and M. Wahde: Fast LIDAR-based Road Detection Using Fully Convolutional Neural Networks. IEEE Intelligent Vehicles Symposium 2017.
30 FusionNet
This method uses stereo information.
96.01 % 94.38 % 95.22 % 96.81 % 5.34 % 3.19 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
31 VGGFCN-6D
This method makes use of Velodyne laser scans.
95.94 % 92.47 % 96.48 % 95.41 % 3.83 % 4.59 % .006 s GPU @ 3.5 Ghz (Python)
32 FuseNet
This method uses stereo information.
95.61 % 95.80 % 95.29 % 95.93 % 5.21 % 4.07 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
33 Up-Conv-Poly code 95.52 % 92.86 % 95.37 % 95.67 % 5.10 % 4.33 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
34 DEEP-DIG 95.45 % 95.41 % 95.49 % 95.41 % 4.96 % 4.59 % 0.14 s GPU @ 3.5 Ghz (Python + C/C++)
J. Muñoz-Bulnes, C. Fernandez, I. Parra, D. Fernández-Llorca and M. Sotelo: Deep Fully Convolutional Networks with Random Data Augmentation for Enhanced Generalization in Road Detection. Workshop on Deep Learning for Autonomous Driving on IEEE 20th International Conference on Intelligent Transportation Systems 2017.
35 ResNetPK
This method makes use of Velodyne laser scans.
95.45 % 92.27 % 96.26 % 94.65 % 4.04 % 5.35 % 0.4s GPU @ 1.5 Ghz (Python)
36 RDSN 95.32 % 91.01 % 94.87 % 95.76 % 5.69 % 4.24 % 0.25 s GPU @ 2.5 Ghz (Python)
37 FCN-GCBs 95.29 % 91.27 % 95.16 % 95.43 % 5.33 % 4.57 % 0.08 s GPU @ 2.5 Ghz (C/C++)
38 SUNet 95.13 % 91.55 % 95.46 % 94.80 % 4.95 % 5.20 % 0.018s
39 RSNetVGG 94.88 % 95.00 % 95.73 % 94.05 % 4.61 % 5.95 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
40 DFN 94.63 % 92.15 % 94.59 % 94.67 % 5.95 % 5.33 % 0.25 s GPU @ >3.5 Ghz (Python)
41 HID-LS
This method makes use of Velodyne laser scans.
94.36 % 91.01 % 94.88 % 93.84 % 5.57 % 6.16 % 0.5 s 4 cores @ 3.0 Ghz (C/C++)
42 DDN 94.17 % 92.70 % 96.73 % 91.74 % 3.41 % 8.26 % 2 s GPU @ 2.5 Ghz (Python + C/C++)
R. Mohan: Deep Deconvolutional Networks for Scene Parsing. 2014.
43 FCN-LC 94.09 % 90.26 % 94.05 % 94.13 % 6.55 % 5.87 % 0.03 s GPU Titan X
C. Mendes, V. Frémont and D. Wolf: Exploiting Fully Convolutional Neural Networks for Fast Road Detection. IEEE Conference on Robotics and Automation (ICRA) 2016.
44 Up-Conv 93.89 % 92.62 % 94.57 % 93.22 % 5.89 % 6.78 % 0.05 s GPU @ 2.5 Ghz (C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
45 HIM 93.55 % 90.38 % 94.18 % 92.92 % 6.31 % 7.08 % 7 s >8 cores @ 2.5 Ghz (Python + C/C++)
D. Munoz, J. Bagnell and M. Hebert: Stacked Hierarchical Labeling. European Conference on Computer Vision (ECCV) 2010.
46 LiDAR-SPHnet
This method makes use of Velodyne laser scans.
93.54 % 89.86 % 93.45 % 93.63 % 7.22 % 6.37 % 0.14 s GPU @ 1.5 Ghz (Matlab)
47 LidarHisto
This method makes use of Velodyne laser scans.
code 93.32 % 93.19 % 95.39 % 91.34 % 4.85 % 8.66 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Chen, J. Yang and H. Kong: Lidar-histogram for fast road and obstacle detection. 2017 IEEE International Conference on Robotics and Automation (ICRA) 2017.
48 StixelNet 93.26 % 87.15 % 90.63 % 96.06 % 10.92 % 3.94 % 1 s GPU @ 2.5 Ghz (C/C++)
D. Levi, N. Garnett and E. Fetaya: StixelNet: A Deep Convolutional Network for Obstacle Detection and Road Segmentation.. 26TH British Machine Vision Conference (BMVC) 2015.
49 HFM 93.12 % 87.10 % 90.58 % 95.82 % 10.96 % 4.18 % 5 s 2 cores @ 2.0 Ghz (C/C++)
50 ResAXN 92.99 % 94.76 % 93.75 % 92.24 % 6.76 % 7.76 % 0.06 s GPU @ 1.5 Ghz (Python)
51 FTP 92.98 % 92.89 % 91.84 % 94.15 % 9.20 % 5.85 % 0.28 s GPU @ 2.5 Ghz (C/C++)
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
52 LWDS 92.81 % 94.66 % 94.41 % 91.25 % 5.93 % 8.75 % 0.07 s GPU @ 2.5 Ghz (Python)
53 MixedCRF
This method makes use of Velodyne laser scans.
92.75 % 90.24 % 94.03 % 91.50 % 6.39 % 8.50 % 6s 1 core @ 2.5 Ghz (Matlab + C/C++)
54 HybridCRF
This method makes use of Velodyne laser scans.
91.95 % 86.44 % 94.01 % 89.98 % 6.30 % 10.02 % 1.5 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, R. Wang, B. Dai, Y. Fang, D. Liu and T. Wu: Hybrid Conditional Random Field based Camera-LIDAR Fusion for Road Detection. Information Sciences. in press.
55 PGM-ARS 91.76 % 84.80 % 88.05 % 95.80 % 14.30 % 4.20 % 0.05 s i74700MQ @ 2.1Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: Fast Pixelwise Road Inference Based on Uniformly Reweighted Belief Propagation . Proc. IEEE Intelligent Vehicles Symposium 2015.
56 TFSeg 91.41 % 93.00 % 91.68 % 91.15 % 9.09 % 8.85 % 0.07 s GPU @ 1.0 Ghz (Python)
57 ProbBoost
This method uses stereo information.
91.36 % 84.92 % 88.18 % 94.78 % 13.97 % 5.22 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: A probabilistic distribution approach for the classification of urban roads in complex environments. Workshop on Modelling, Estimation, Perception and Control of All Terrain Mobile Robots on IEEE International Conference on Robotics and Automation (ICRA) 2014.
58 NNP
This method uses stereo information.
91.34 % 88.65 % 91.07 % 91.60 % 9.87 % 8.40 % 5 s 4 cores @ 2.5 Ghz (Matlab)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
59 FRS_SP 90.96 % 84.63 % 87.86 % 94.29 % 14.32 % 5.71 % 0.21 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
60 SRF 90.77 % 92.44 % 89.35 % 92.23 % 12.08 % 7.77 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, D. Zhao and T. Wu: Monocular Road Detection Using Structured Random Forest. Int J Adv Robot Syst 2016.
61 RES3D-Velo
This method makes use of Velodyne laser scans.
90.60 % 85.38 % 85.96 % 95.78 % 17.20 % 4.22 % 0.36 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Wolf and C. Stiller: Road Terrain Detection: Avoiding Common Obstacle Detection Assumptions Using Sensor Fusion. Intelligent Vehicles Symposium (IV) 2014.
62 CB 90.55 % 85.40 % 92.75 % 88.45 % 7.60 % 11.55 % 2 s 1 core @ 3.4 Ghz (Python) + GPU
C. Mendes, V. Frémont and D. Wolf: Vision-Based Road Detection using Contextual Blocks. 2015.
63 MAP 89.97 % 92.14 % 87.47 % 92.62 % 14.58 % 7.38 % 0.28s GPU
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
64 DVFCN 89.95 % 93.93 % 89.74 % 90.17 % 11.33 % 9.83 % 0.07 s GPU @ 2.5 Ghz (Python)
65 SPRAY 89.69 % 93.84 % 89.13 % 90.25 % 12.10 % 9.75 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
66 ARSL-AMI 89.56 % 82.82 % 85.87 % 93.59 % 16.93 % 6.41 % 0.05 s 4 cores @ 2.5 Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: CRF-based semantic labeling in miniaturized road scenes . Proc. IEEE Intelligent Transportation Systems 2014.
67 FusedCRF
This method makes use of Velodyne laser scans.
89.51 % 83.53 % 86.64 % 92.58 % 15.69 % 7.42 % 2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, T. Hu and T. Wu: CRF based Road Detection with Multi-Sensor Fusion. Intelligent Vehicles Symposium (IV) 2015.
68 BM
This method uses stereo information.
89.41 % 80.61 % 83.43 % 96.30 % 21.02 % 3.70 % 2 s 2 cores @ 2.5 Ghz (Matlab)
B. WANG, V. Fremont and S. Rodriguez Florez: Color-based Road Detection and its Evaluation on the KITTI Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application, IEEE Intelligent Vehicles Symposium 2014.
69 HistonBoost
This method uses stereo information.
88.73 % 81.57 % 84.49 % 93.42 % 18.85 % 6.58 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: Comprehensive Performance Analysis of Road Detection Algorithms Using the Common Urban Kitti-Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application on IEEE Intelligent Vehicles Symposium (IV) 2014.
70 SegNet 88.59 % 83.54 % 88.35 % 88.84 % 12.88 % 11.16 % 0.01 s 8 cores @ 2.5 Ghz (C/C++)
71 geo+gpr+crf
This method uses stereo information.
88.20 % 82.33 % 85.32 % 91.27 % 17.26 % 8.73 % 30 s 1 core @ 2.0 Ghz (C/C++)
Z. Xiao, B. Dai, H. Li, T. Wu, X. Xu, Y. Zeng and T. Chen: Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion. International Journal of Advanced Robotic Systems 2017.
72 GRES3D+VELO
This method makes use of Velodyne laser scans.
88.19 % 88.65 % 83.98 % 92.85 % 19.48 % 7.15 % 60 ms 4 cores @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
73 SCRFFPFHGSP
This method uses stereo information.
87.96 % 83.16 % 90.01 % 86.01 % 10.50 % 13.99 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
74 Pos-ex code 87.75 % 90.75 % 82.54 % 93.67 % 21.78 % 6.33 % 120 ms GPU(K20) @ 0.7 Ghz (Matlab + Caffe)
75 GRES3D+SELAS
This method uses stereo information.
87.57 % 90.52 % 85.92 % 89.28 % 16.08 % 10.72 % 110 ms 4 core @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
76 Strait 86.36 % 91.98 % 85.75 % 86.97 % 15.89 % 13.03 % 69 ms GPU @ K20
77 CN 86.21 % 84.40 % 82.85 % 89.86 % 20.45 % 10.14 % 2 s 1 core @ 2.5 Ghz (C/C++)
J. Alvarez, T. Gevers, Y. LeCun and A. Lopez: Road Scene Segmentation from a Single Image. ECCV 2012 2012.
78 LWD 86.14 % 88.31 % 86.99 % 85.31 % 14.03 % 14.69 % 0.07 s GPU @ 2.5 Ghz (Python)
79 SP-SS 85.07 % 79.86 % 85.97 % 84.20 % 15.11 % 15.80 % 0.01 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
80 CNN_LSTM 84.98 % 83.43 % 90.34 % 80.22 % 9.43 % 19.78 % 54 ms GPU @ 2.0 Ghz (Python + C/C++)
81 RES3D-Stereo
This method uses stereo information.
83.62 % 85.74 % 79.81 % 87.81 % 24.42 % 12.19 % 0.7 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Gomes and D. Wolf: Road estimation with sparse 3D points from stereo data. Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on 2014.
82 SPlane
This method uses stereo information.
82.28 % 82.83 % 76.85 % 88.53 % 29.32 % 11.47 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
83 SPlane + BL
This method uses stereo information.
82.04 % 85.56 % 75.11 % 90.39 % 32.93 % 9.61 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
84 ANN
This method uses stereo information.
80.95 % 68.36 % 69.95 % 96.05 % 45.35 % 3.95 % 3 s 1 core @ 3.0 Ghz (C/C++)
G. Vitor, D. Lima, A. Victorino and J. Ferreira: A 2D/3D Vision Based Approach Applied to Road Detection in Urban Environments. Intelligent Vehicles Symposium (IV), 2013 IEEE 2013.
85 LKW 75.48 % 78.73 % 65.97 % 88.18 % 50.00 % 11.82 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
86 VAP 71.83 % 60.64 % 62.48 % 84.48 % 55.76 % 15.52 % 1 s 1 core @ 2.5 Ghz (Matlab)
ERROR: Wrong syntax in BIBTEX file.
Table as LaTeX | Only published Methods

UU_ROAD


Method Setting Code MaxF AP PRE REC FPR FNR Runtime Environment
1 iDST-VT 96.06 % 92.89 % 95.82 % 96.30 % 1.37 % 3.70 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
2 UNV 95.71 % 90.32 % 95.13 % 96.30 % 1.61 % 3.70 % 1.2 s GPU @ 3.0 Ghz (Python + C/C++)
3 DFFA 95.57 % 89.21 % 95.67 % 95.47 % 1.41 % 4.53 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
4 KRS 95.55 % 92.90 % 95.81 % 95.30 % 1.36 % 4.70 % 0.3 s GPU @ 2.5 Ghz (Python)
5 KRSF 95.53 % 92.96 % 96.25 % 94.83 % 1.21 % 5.17 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
6 NF2CNN
This method makes use of Velodyne laser scans.
95.47 % 86.98 % 93.22 % 97.84 % 2.32 % 2.16 % .006 s GPU @ 3.5 Ghz (Python)
7 YhY code 95.39 % 88.50 % 94.89 % 95.90 % 1.68 % 4.10 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
8 DCCN
This method makes use of Velodyne laser scans.
94.89 % 92.27 % 94.98 % 94.80 % 1.63 % 5.20 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
9 RSNet 94.84 % 91.65 % 94.56 % 95.13 % 1.78 % 4.87 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
10 KRS 94.60 % 92.83 % 94.96 % 94.25 % 1.63 % 5.75 % 1 s GPU @ 2.5 Ghz (Python)
11 LidCamNet
This method makes use of Velodyne laser scans.
94.54 % 92.74 % 94.64 % 94.45 % 1.74 % 5.55 % 0.15 s GPU @ 2.5 Ghz (Python)
12 WSLGAN 94.54 % 87.70 % 94.01 % 95.09 % 1.97 % 4.91 % 800ms GPU @ 1.5 Ghz (Python)
13 SSLGAN 94.40 % 87.84 % 94.17 % 94.63 % 1.91 % 5.37 % 700ms GPU @ 1.5 Ghz (Python)
14 baseline 94.14 % 92.15 % 93.69 % 94.60 % 2.08 % 5.40 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
15 MVnet
This method makes use of Velodyne laser scans.
94.07 % 89.03 % 95.48 % 92.71 % 1.43 % 7.29 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
16 RSNet2 93.98 % 91.88 % 93.81 % 94.15 % 2.03 % 5.85 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
17 RSNet- 93.90 % 91.69 % 92.94 % 94.88 % 2.35 % 5.12 % 0.07 s GPU @ 2.5 Ghz (Python)
18 MMN 93.87 % 91.23 % 93.26 % 94.48 % 2.22 % 5.52 % 0.1 s GPU @ 2.5 Ghz (C/C++)
19 CoDNN 93.81 % 92.41 % 94.20 % 93.43 % 1.88 % 6.57 % 0.01 s 1 core @ 2.5 Ghz (Python + C/C++)
20 FNETMS 93.71 % 91.79 % 94.08 % 93.34 % 1.91 % 6.66 % 0.04 s GPU @ 2.0 Ghz (Python + C/C++)
21 MultiNet code 93.69 % 92.55 % 94.24 % 93.14 % 1.85 % 6.86 % 0.17 s GPU @ 2.5 Ghz (Python + C/C++)
M. Teichmann, M. Weber, J. Zoellner, R. Cipolla and R. Urtasun: MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving. CoRR 2016.
22 IDA-Fusion
This method makes use of Velodyne laser scans.
93.63 % 86.07 % 92.22 % 95.08 % 2.61 % 4.92 % 0.1 s 4 cores @ 3.5 Ghz (C/C++)
23 wt 93.58 % 92.14 % 93.34 % 93.83 % 2.18 % 6.17 % 0.1 s GPU @ 1.0 Ghz (Python)
24 StixelNet II 93.40 % 85.01 % 91.05 % 95.87 % 3.07 % 4.13 % 1.2 s 1 core @ 3.0 Ghz (Matlab + C/C++)
N. Garnett, S. Silberstein, S. Oron, E. Fetaya, U. Verner, A. Ayash, V. Goldner, R. Cohen, K. Horn and D. Levi: Real-time category-based and general obstacle detection for autonomous driving. 5th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving (CVRSUAD'17, IEEE-ICCV 2017 Workshop) 2017.
25 RBNet 93.21 % 89.18 % 92.81 % 93.60 % 2.36 % 6.40 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
26 FDN 93.10 % 92.24 % 93.44 % 92.77 % 2.12 % 7.23 % 0.2 s GPU 1 core @ 2.5 Ghz (Python)
27 FuseNet
This method uses stereo information.
92.97 % 92.47 % 93.44 % 92.51 % 2.12 % 7.49 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
28 FusionNet
This method uses stereo information.
92.89 % 90.69 % 92.75 % 93.03 % 2.37 % 6.97 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
29 MBN 92.89 % 83.53 % 89.43 % 96.63 % 3.72 % 3.37 % 0.16 s GPU @ 2.5 Ghz (Python)
30 Up-Conv-Poly code 92.65 % 89.20 % 92.85 % 92.45 % 2.32 % 7.55 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
31 ResNetPK
This method makes use of Velodyne laser scans.
92.56 % 86.93 % 93.16 % 91.96 % 2.20 % 8.04 % 0.4s GPU @ 1.5 Ghz (Python)
32 TDCac1 CNN 92.30 % 86.21 % 92.37 % 92.23 % 2.48 % 7.77 % .093 s 1 core @ 1.0 Ghz (C/C++)
33 LoDNN
This method makes use of Velodyne laser scans.
92.29 % 90.35 % 90.81 % 93.81 % 3.09 % 6.19 % 18 ms GPU @ 2.5 Ghz (Torch)
L. Caltagirone, S. Scheidegger, L. Svensson and M. Wahde: Fast LIDAR-based Road Detection Using Fully Convolutional Neural Networks. IEEE Intelligent Vehicles Symposium 2017.
34 FCN-GCBs 92.10 % 83.69 % 89.61 % 94.73 % 3.58 % 5.27 % 0.08 s GPU @ 2.5 Ghz (C/C++)
35 RDSN 91.98 % 85.64 % 91.45 % 92.53 % 2.82 % 7.47 % 0.25 s GPU @ 2.5 Ghz (Python)
36 Up-Conv 91.89 % 89.44 % 92.59 % 91.20 % 2.38 % 8.80 % 0.05 s GPU @ 2.5 Ghz (C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
37 DDN 91.76 % 86.84 % 93.06 % 90.50 % 2.20 % 9.50 % 2 s GPU @ 2.5 Ghz (Python + C/C++)
R. Mohan: Deep Deconvolutional Networks for Scene Parsing. 2014.
38 RSNetVGG 91.72 % 91.52 % 92.62 % 90.84 % 2.36 % 9.16 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
39 DFN 91.70 % 87.72 % 91.51 % 91.89 % 2.78 % 8.11 % 0.25 s GPU @ >3.5 Ghz (Python)
40 DEEP-DIG 91.27 % 91.77 % 91.32 % 91.22 % 2.82 % 8.78 % 0.14 s GPU @ 3.5 Ghz (Python + C/C++)
J. Muñoz-Bulnes, C. Fernandez, I. Parra, D. Fernández-Llorca and M. Sotelo: Deep Fully Convolutional Networks with Random Data Augmentation for Enhanced Generalization in Road Detection. Workshop on Deep Learning for Autonomous Driving on IEEE 20th International Conference on Intelligent Transportation Systems 2017.
41 VGGFCN-6D
This method makes use of Velodyne laser scans.
91.25 % 84.38 % 90.36 % 92.16 % 3.20 % 7.84 % .006 s GPU @ 3.5 Ghz (Python)
42 SUNet 91.10 % 81.62 % 87.32 % 95.22 % 4.50 % 4.78 % 0.018s
43 VD 90.03 % 78.89 % 90.96 % 89.11 % 2.89 % 10.89 % 4 s 1 core @ 3.0 Ghz (Matlab + C/C++)
44 FTP 89.62 % 88.93 % 89.10 % 90.14 % 3.59 % 9.86 % 0.28 s GPU @ 2.5 Ghz (C/C++)
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
45 HFM 89.20 % 80.48 % 86.07 % 92.56 % 4.88 % 7.44 % 5 s 2 cores @ 2.0 Ghz (C/C++)
46 HID-LS
This method makes use of Velodyne laser scans.
89.10 % 80.53 % 86.13 % 92.29 % 4.84 % 7.71 % 0.5 s 4 cores @ 3.0 Ghz (C/C++)
47 LiDAR-SPHnet
This method makes use of Velodyne laser scans.
88.98 % 80.48 % 86.07 % 92.09 % 4.86 % 7.91 % 0.14 s GPU @ 1.5 Ghz (Matlab)
48 HybridCRF
This method makes use of Velodyne laser scans.
88.53 % 80.79 % 86.41 % 90.76 % 4.65 % 9.24 % 1.5 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, R. Wang, B. Dai, Y. Fang, D. Liu and T. Wu: Hybrid Conditional Random Field based Camera-LIDAR Fusion for Road Detection. Information Sciences. in press.
49 TFSeg 87.42 % 88.58 % 87.28 % 87.56 % 4.16 % 12.44 % 0.07 s GPU @ 1.0 Ghz (Python)
50 LidarHisto
This method makes use of Velodyne laser scans.
code 86.55 % 81.13 % 90.71 % 82.75 % 2.76 % 17.25 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Chen, J. Yang and H. Kong: Lidar-histogram for fast road and obstacle detection. 2017 IEEE International Conference on Robotics and Automation (ICRA) 2017.
51 FCN-LC 86.27 % 75.37 % 86.65 % 85.89 % 4.31 % 14.11 % 0.03 s GPU Titan X
C. Mendes, V. Frémont and D. Wolf: Exploiting Fully Convolutional Neural Networks for Fast Road Detection. IEEE Conference on Robotics and Automation (ICRA) 2016.
52 CB 86.13 % 75.21 % 86.47 % 85.80 % 4.38 % 14.20 % 2 s 1 core @ 3.4 Ghz (Python) + GPU
C. Mendes, V. Frémont and D. Wolf: Vision-Based Road Detection using Contextual Blocks. 2015.
53 StixelNet 86.06 % 72.05 % 82.61 % 89.82 % 6.16 % 10.18 % 1 s GPU @ 2.5 Ghz (C/C++)
D. Levi, N. Garnett and E. Fetaya: StixelNet: A Deep Convolutional Network for Obstacle Detection and Road Segmentation.. 26TH British Machine Vision Conference (BMVC) 2015.
54 HIM 85.76 % 76.18 % 87.65 % 83.95 % 3.86 % 16.05 % 7 s >8 cores @ 2.5 Ghz (Python + C/C++)
D. Munoz, J. Bagnell and M. Hebert: Stacked Hierarchical Labeling. European Conference on Computer Vision (ECCV) 2010.
55 MixedCRF
This method makes use of Velodyne laser scans.
85.69 % 75.12 % 80.17 % 92.02 % 7.42 % 7.98 % 6s 1 core @ 2.5 Ghz (Matlab + C/C++)
56 NNP
This method uses stereo information.
85.55 % 76.90 % 85.36 % 85.75 % 4.79 % 14.25 % 5 s 4 cores @ 2.5 Ghz (Matlab)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
57 DVFCN 85.37 % 89.13 % 85.44 % 85.30 % 4.74 % 14.70 % 0.07 s GPU @ 2.5 Ghz (Python)
58 LWD 84.96 % 87.51 % 82.18 % 87.93 % 6.21 % 12.07 % 0.07 s GPU @ 2.5 Ghz (Python)
59 LWDS 84.50 % 83.24 % 87.46 % 81.72 % 3.82 % 18.28 % 0.07 s GPU @ 2.5 Ghz (Python)
60 FusedCRF
This method makes use of Velodyne laser scans.
84.49 % 72.35 % 77.13 % 93.40 % 9.02 % 6.60 % 2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, T. Hu and T. Wu: CRF based Road Detection with Multi-Sensor Fusion. Intelligent Vehicles Symposium (IV) 2015.
61 MAP 84.44 % 87.17 % 83.66 % 85.23 % 5.42 % 14.77 % 0.28s GPU
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
62 GRES3D+VELO
This method makes use of Velodyne laser scans.
84.14 % 80.20 % 80.57 % 88.03 % 6.92 % 11.97 % 60 ms 4 cores @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
63 ResAXN 83.68 % 87.44 % 84.04 % 83.33 % 5.16 % 16.67 % 0.06 s GPU @ 1.5 Ghz (Python)
64 RES3D-Velo
This method makes use of Velodyne laser scans.
83.63 % 72.58 % 77.38 % 90.97 % 8.67 % 9.03 % 0.36 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Wolf and C. Stiller: Road Terrain Detection: Avoiding Common Obstacle Detection Assumptions Using Sensor Fusion. Intelligent Vehicles Symposium (IV) 2014.
65 SPRAY 82.71 % 87.19 % 82.16 % 83.26 % 5.89 % 16.74 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
66 GRES3D+SELAS
This method uses stereo information.
82.70 % 83.95 % 78.54 % 87.32 % 7.77 % 12.68 % 110 ms 4 core @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
67 geo+gpr+crf
This method uses stereo information.
81.00 % 69.74 % 79.78 % 82.27 % 6.79 % 17.73 % 30 s 1 core @ 2.0 Ghz (C/C++)
Z. Xiao, B. Dai, H. Li, T. Wu, X. Xu, Y. Zeng and T. Chen: Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion. International Journal of Advanced Robotic Systems 2017.
68 SCRFFPFHGSP
This method uses stereo information.
80.78 % 70.80 % 81.07 % 80.50 % 6.13 % 19.50 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
69 ProbBoost
This method uses stereo information.
80.76 % 68.70 % 85.25 % 76.72 % 4.33 % 23.28 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: A probabilistic distribution approach for the classification of urban roads in complex environments. Workshop on Modelling, Estimation, Perception and Control of All Terrain Mobile Robots on IEEE International Conference on Robotics and Automation (ICRA) 2014.
70 FRS_SP 80.02 % 67.93 % 77.56 % 82.64 % 7.79 % 17.36 % 0.21 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
71 PGM-ARS 79.94 % 67.77 % 77.37 % 82.67 % 7.88 % 17.33 % 0.05 s i74700MQ @ 2.1Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: Fast Pixelwise Road Inference Based on Uniformly Reweighted Belief Propagation . Proc. IEEE Intelligent Vehicles Symposium 2015.
72 Strait 79.75 % 84.94 % 78.52 % 81.02 % 7.22 % 18.98 % 69 ms GPU @ K20
73 RES3D-Stereo
This method uses stereo information.
78.75 % 73.60 % 77.63 % 79.90 % 7.50 % 20.10 % 0.7 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Gomes and D. Wolf: Road estimation with sparse 3D points from stereo data. Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on 2014.
74 SP-SS 78.47 % 65.18 % 74.20 % 83.25 % 9.43 % 16.75 % 0.01 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
75 BM
This method uses stereo information.
78.43 % 62.46 % 70.87 % 87.80 % 11.76 % 12.20 % 2 s 2 cores @ 2.5 Ghz (Matlab)
B. WANG, V. Fremont and S. Rodriguez Florez: Color-based Road Detection and its Evaluation on the KITTI Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application, IEEE Intelligent Vehicles Symposium 2014.
76 SegNet 77.23 % 69.23 % 82.29 % 72.76 % 5.10 % 27.24 % 0.01 s 8 cores @ 2.5 Ghz (C/C++)
77 CNN_LSTM 76.28 % 65.25 % 80.51 % 72.47 % 5.72 % 27.53 % 54 ms GPU @ 2.0 Ghz (Python + C/C++)
78 SRF 76.07 % 79.97 % 71.47 % 81.31 % 10.57 % 18.69 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, D. Zhao and T. Wu: Monocular Road Detection Using Structured Random Forest. Int J Adv Robot Syst 2016.
79 Pos-ex code 74.60 % 79.21 % 70.52 % 79.19 % 10.79 % 20.81 % 120 ms GPU(K20) @ 0.7 Ghz (Matlab + Caffe)
80 HistonBoost
This method uses stereo information.
74.19 % 63.01 % 77.43 % 71.22 % 6.77 % 28.78 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: Comprehensive Performance Analysis of Road Detection Algorithms Using the Common Urban Kitti-Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application on IEEE Intelligent Vehicles Symposium (IV) 2014.
81 SPlane + BL
This method uses stereo information.
74.02 % 79.61 % 65.15 % 85.68 % 14.93 % 14.32 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
82 SPlane
This method uses stereo information.
73.30 % 69.11 % 65.39 % 83.38 % 14.38 % 16.62 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
83 CN 72.25 % 66.61 % 71.96 % 72.54 % 9.21 % 27.46 % 2 s 1 core @ 2.5 Ghz (C/C++)
J. Alvarez, T. Gevers, Y. LeCun and A. Lopez: Road Scene Segmentation from a Single Image. ECCV 2012 2012.
84 ARSL-AMI 70.33 % 61.97 % 83.33 % 60.84 % 3.97 % 39.16 % 0.05 s 4 cores @ 2.5 Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: CRF-based semantic labeling in miniaturized road scenes . Proc. IEEE Intelligent Transportation Systems 2014.
85 LKW 69.65 % 74.02 % 65.45 % 74.42 % 12.80 % 25.58 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
86 VAP 54.62 % 37.65 % 38.96 % 91.32 % 46.62 % 8.68 % 1 s 1 core @ 2.5 Ghz (Matlab)
ERROR: Wrong syntax in BIBTEX file.
87 ANN
This method uses stereo information.
54.07 % 36.61 % 39.28 % 86.69 % 43.67 % 13.31 % 3 s 1 core @ 3.0 Ghz (C/C++)
G. Vitor, D. Lima, A. Victorino and J. Ferreira: A 2D/3D Vision Based Approach Applied to Road Detection in Urban Environments. Intelligent Vehicles Symposium (IV), 2013 IEEE 2013.
Table as LaTeX | Only published Methods

URBAN_ROAD


Method Setting Code MaxF AP PRE REC FPR FNR Runtime Environment
1 iDST-VT 96.76 % 93.90 % 96.63 % 96.89 % 1.86 % 3.11 % 1 s GPU @ 2.5 Ghz (Python + C/C++)
2 UNV 96.74 % 92.39 % 96.88 % 96.60 % 1.71 % 3.40 % 1.2 s GPU @ 3.0 Ghz (Python + C/C++)
3 NF2CNN
This method makes use of Velodyne laser scans.
96.70 % 89.93 % 95.37 % 98.07 % 2.62 % 1.93 % .006 s GPU @ 3.5 Ghz (Python)
4 KRSF 96.50 % 94.01 % 96.74 % 96.27 % 1.78 % 3.73 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
5 YhY code 96.44 % 90.43 % 95.92 % 96.97 % 2.27 % 3.03 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
6 KRS 96.41 % 93.96 % 96.43 % 96.38 % 1.96 % 3.62 % 0.3 s GPU @ 2.5 Ghz (Python)
7 DFFA 96.35 % 90.52 % 96.02 % 96.69 % 2.21 % 3.31 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
8 LidCamNet
This method makes use of Velodyne laser scans.
96.03 % 93.93 % 96.23 % 95.83 % 2.07 % 4.17 % 0.15 s GPU @ 2.5 Ghz (Python)
9 DCCN
This method makes use of Velodyne laser scans.
95.93 % 93.32 % 95.52 % 96.35 % 2.49 % 3.65 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
10 RSNet 95.86 % 93.21 % 95.68 % 96.05 % 2.39 % 3.95 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
11 MVnet
This method makes use of Velodyne laser scans.
95.83 % 91.67 % 97.29 % 94.41 % 1.45 % 5.59 % 0.04 s 1 core @ 2.5 Ghz (Python + C/C++)
12 WSLGAN 95.70 % 90.17 % 95.64 % 95.77 % 2.40 % 4.23 % 800ms GPU @ 1.5 Ghz (Python)
13 KRS 95.64 % 93.89 % 95.79 % 95.48 % 2.31 % 4.52 % 1 s GPU @ 2.5 Ghz (Python)
14 baseline 95.58 % 93.51 % 95.19 % 95.97 % 2.67 % 4.03 % 0.05 s 1 core @ 2.5 Ghz (Python + C/C++)
15 SSLGAN 95.53 % 90.35 % 95.84 % 95.24 % 2.28 % 4.76 % 700ms GPU @ 1.5 Ghz (Python)
16 RSNet2 95.35 % 93.32 % 95.20 % 95.49 % 2.65 % 4.51 % 0.07 s GPU @ 2.5 Ghz (Python + C/C++)
17 RSNet- 95.29 % 93.37 % 94.96 % 95.62 % 2.80 % 4.38 % 0.07 s GPU @ 2.5 Ghz (Python)
18 IDA-Fusion
This method makes use of Velodyne laser scans.
95.22 % 89.31 % 94.69 % 95.76 % 2.96 % 4.24 % 0.1 s 4 cores @ 3.5 Ghz (C/C++)
19 MMN 95.12 % 92.99 % 94.82 % 95.42 % 2.87 % 4.58 % 0.1 s GPU @ 2.5 Ghz (C/C++)
20 CoDNN 95.06 % 93.58 % 95.24 % 94.88 % 2.61 % 5.12 % 0.01 s 1 core @ 2.5 Ghz (Python + C/C++)
21 FNETMS 94.99 % 93.18 % 94.90 % 95.09 % 2.82 % 4.91 % 0.04 s GPU @ 2.0 Ghz (Python + C/C++)
22 RBNet 94.97 % 91.49 % 94.94 % 95.01 % 2.79 % 4.99 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
23 StixelNet II 94.88 % 87.75 % 92.97 % 96.87 % 4.04 % 3.13 % 1.2 s 1 core @ 3.0 Ghz (Matlab + C/C++)
N. Garnett, S. Silberstein, S. Oron, E. Fetaya, U. Verner, A. Ayash, V. Goldner, R. Cohen, K. Horn and D. Levi: Real-time category-based and general obstacle detection for autonomous driving. 5th Workshop on Computer Vision for Road Scene Understanding and Autonomous Driving (CVRSUAD'17, IEEE-ICCV 2017 Workshop) 2017.
24 MultiNet code 94.88 % 93.71 % 94.84 % 94.91 % 2.85 % 5.09 % 0.17 s GPU @ 2.5 Ghz (Python + C/C++)
M. Teichmann, M. Weber, J. Zoellner, R. Cipolla and R. Urtasun: MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving. CoRR 2016.
25 TDCac1 CNN 94.81 % 89.82 % 95.25 % 94.38 % 2.59 % 5.62 % .093 s 1 core @ 1.0 Ghz (C/C++)
26 FDN 94.75 % 93.63 % 94.65 % 94.86 % 2.96 % 5.14 % 0.2 s GPU 1 core @ 2.5 Ghz (Python)
27 FusionNet
This method uses stereo information.
94.67 % 92.54 % 94.73 % 94.61 % 2.90 % 5.39 % 0.3 s GPU @ 2.5 Ghz (Python + C/C++)
28 wt 94.63 % 93.56 % 94.38 % 94.88 % 3.11 % 5.12 % 0.1 s GPU @ 1.0 Ghz (Python)
29 MBN 94.63 % 87.37 % 92.55 % 96.80 % 4.29 % 3.20 % 0.16 s GPU @ 2.5 Ghz (Python)
30 FuseNet
This method uses stereo information.
94.41 % 93.77 % 94.70 % 94.11 % 2.90 % 5.89 % 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
31 ResNetPK
This method makes use of Velodyne laser scans.
94.25 % 89.66 % 95.07 % 93.45 % 2.67 % 6.55 % 0.4s GPU @ 1.5 Ghz (Python)
32 FCN-GCBs 94.08 % 87.66 % 92.87 % 95.32 % 4.03 % 4.68 % 0.08 s GPU @ 2.5 Ghz (C/C++)
33 LoDNN
This method makes use of Velodyne laser scans.
94.07 % 92.03 % 92.81 % 95.37 % 4.07 % 4.63 % 18 ms GPU @ 2.5 Ghz (Torch)
L. Caltagirone, S. Scheidegger, L. Svensson and M. Wahde: Fast LIDAR-based Road Detection Using Fully Convolutional Neural Networks. IEEE Intelligent Vehicles Symposium 2017.
34 DEEP-DIG 93.98 % 93.65 % 94.26 % 93.69 % 3.14 % 6.31 % 0.14 s GPU @ 3.5 Ghz (Python + C/C++)
J. Muñoz-Bulnes, C. Fernandez, I. Parra, D. Fernández-Llorca and M. Sotelo: Deep Fully Convolutional Networks with Random Data Augmentation for Enhanced Generalization in Road Detection. Workshop on Deep Learning for Autonomous Driving on IEEE 20th International Conference on Intelligent Transportation Systems 2017.
35 Up-Conv-Poly code 93.83 % 90.47 % 94.00 % 93.67 % 3.29 % 6.33 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
36 RDSN 93.75 % 88.33 % 93.55 % 93.96 % 3.57 % 6.04 % 0.25 s GPU @ 2.5 Ghz (Python)
37 VGGFCN-6D
This method makes use of Velodyne laser scans.
93.72 % 88.61 % 93.92 % 93.52 % 3.33 % 6.48 % .006 s GPU @ 3.5 Ghz (Python)
38 DDN 93.43 % 89.67 % 95.09 % 91.82 % 2.61 % 8.18 % 2 s GPU @ 2.5 Ghz (Python + C/C++)
R. Mohan: Deep Deconvolutional Networks for Scene Parsing. 2014.
39 SUNet 93.38 % 87.18 % 92.35 % 94.44 % 4.31 % 5.56 % 0.018s
40 RSNetVGG 93.24 % 93.10 % 94.68 % 91.84 % 2.85 % 8.16 % 0.09 s GPU @ 2.5 Ghz (Python + C/C++)
41 DFN 93.23 % 89.59 % 93.11 % 93.35 % 3.81 % 6.65 % 0.25 s GPU @ >3.5 Ghz (Python)
42 Up-Conv 92.39 % 90.24 % 93.03 % 91.76 % 3.79 % 8.24 % 0.05 s GPU @ 2.5 Ghz (C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
43 HID-LS
This method makes use of Velodyne laser scans.
92.36 % 85.83 % 90.86 % 93.90 % 5.21 % 6.10 % 0.5 s 4 cores @ 3.0 Ghz (C/C++)
44 LiDAR-SPHnet
This method makes use of Velodyne laser scans.
91.79 % 84.76 % 89.68 % 94.00 % 5.96 % 6.00 % 0.14 s GPU @ 1.5 Ghz (Matlab)
45 FTP 91.61 % 90.96 % 91.04 % 92.20 % 5.00 % 7.80 % 0.28 s GPU @ 2.5 Ghz (C/C++)
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
46 HFM 90.88 % 83.10 % 87.86 % 94.12 % 7.16 % 5.88 % 5 s 2 cores @ 2.0 Ghz (C/C++)
47 HybridCRF
This method makes use of Velodyne laser scans.
90.81 % 86.01 % 91.05 % 90.57 % 4.90 % 9.43 % 1.5 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, R. Wang, B. Dai, Y. Fang, D. Liu and T. Wu: Hybrid Conditional Random Field based Camera-LIDAR Fusion for Road Detection. Information Sciences. in press.
48 FCN-LC 90.79 % 85.83 % 90.87 % 90.72 % 5.02 % 9.28 % 0.03 s GPU Titan X
C. Mendes, V. Frémont and D. Wolf: Exploiting Fully Convolutional Neural Networks for Fast Road Detection. IEEE Conference on Robotics and Automation (ICRA) 2016.
49 LidarHisto
This method makes use of Velodyne laser scans.
code 90.67 % 84.79 % 93.06 % 88.41 % 3.63 % 11.59 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
L. Chen, J. Yang and H. Kong: Lidar-histogram for fast road and obstacle detection. 2017 IEEE International Conference on Robotics and Automation (ICRA) 2017.
50 HIM 90.64 % 81.42 % 91.62 % 89.68 % 4.52 % 10.32 % 7 s >8 cores @ 2.5 Ghz (Python + C/C++)
D. Munoz, J. Bagnell and M. Hebert: Stacked Hierarchical Labeling. European Conference on Computer Vision (ECCV) 2010.
51 MixedCRF
This method makes use of Velodyne laser scans.
90.59 % 84.24 % 89.11 % 92.13 % 6.20 % 7.87 % 6s 1 core @ 2.5 Ghz (Matlab + C/C++)
52 LWDS 89.83 % 87.04 % 91.61 % 88.12 % 4.45 % 11.88 % 0.07 s GPU @ 2.5 Ghz (Python)
53 NNP
This method uses stereo information.
89.68 % 86.50 % 89.67 % 89.68 % 5.69 % 10.32 % 5 s 4 cores @ 2.5 Ghz (Matlab)
X. Chen, K. Kundu, Y. Zhu, A. Berneshawi, H. Ma, S. Fidler and R. Urtasun: 3D Object Proposals for Accurate Object Class Detection. NIPS 2015.
54 TFSeg 89.65 % 89.24 % 88.79 % 90.52 % 6.30 % 9.48 % 0.07 s GPU @ 1.0 Ghz (Python)
55 ResAXN 89.39 % 91.91 % 90.84 % 87.98 % 4.89 % 12.02 % 0.06 s GPU @ 1.5 Ghz (Python)
56 StixelNet 89.12 % 81.23 % 85.80 % 92.71 % 8.45 % 7.29 % 1 s GPU @ 2.5 Ghz (C/C++)
D. Levi, N. Garnett and E. Fetaya: StixelNet: A Deep Convolutional Network for Obstacle Detection and Road Segmentation.. 26TH British Machine Vision Conference (BMVC) 2015.
57 CB 88.97 % 79.69 % 89.50 % 88.44 % 5.71 % 11.56 % 2 s 1 core @ 3.4 Ghz (Python) + GPU
C. Mendes, V. Frémont and D. Wolf: Vision-Based Road Detection using Contextual Blocks. 2015.
58 DVFCN 88.34 % 91.70 % 88.51 % 88.17 % 6.30 % 11.83 % 0.07 s GPU @ 2.5 Ghz (Python)
59 FusedCRF
This method makes use of Velodyne laser scans.
88.25 % 79.24 % 83.62 % 93.44 % 10.08 % 6.56 % 2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, T. Hu and T. Wu: CRF based Road Detection with Multi-Sensor Fusion. Intelligent Vehicles Symposium (IV) 2015.
60 MAP 87.80 % 89.96 % 86.01 % 89.66 % 8.04 % 10.34 % 0.28s GPU
A. Laddha, M. Kocamaz, L. Navarro-Serment and M. Hebert: Map-Supervised Road Detection. IEEE Intelligent Vehicles Symposium Proceedings 2016.
61 ProbBoost
This method uses stereo information.
87.78 % 77.30 % 86.59 % 89.01 % 7.60 % 10.99 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: A probabilistic distribution approach for the classification of urban roads in complex environments. Workshop on Modelling, Estimation, Perception and Control of All Terrain Mobile Robots on IEEE International Conference on Robotics and Automation (ICRA) 2014.
62 SPRAY 87.09 % 91.12 % 87.10 % 87.08 % 7.10 % 12.92 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
63 RES3D-Velo
This method makes use of Velodyne laser scans.
86.58 % 78.34 % 82.63 % 90.92 % 10.53 % 9.08 % 0.36 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Wolf and C. Stiller: Road Terrain Detection: Avoiding Common Obstacle Detection Assumptions Using Sensor Fusion. Intelligent Vehicles Symposium (IV) 2014.
64 GRES3D+VELO
This method makes use of Velodyne laser scans.
86.07 % 84.34 % 82.16 % 90.38 % 10.81 % 9.62 % 60 ms 4 cores @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
65 FRS_SP 85.97 % 77.81 % 82.04 % 90.31 % 10.89 % 9.69 % 0.21 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
66 LWD 85.74 % 84.86 % 83.76 % 87.81 % 9.38 % 12.19 % 0.07 s GPU @ 2.5 Ghz (Python)
67 PGM-ARS 85.69 % 73.83 % 82.34 % 89.33 % 10.56 % 10.67 % 0.05 s i74700MQ @ 2.1Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: Fast Pixelwise Road Inference Based on Uniformly Reweighted Belief Propagation . Proc. IEEE Intelligent Vehicles Symposium 2015.
68 geo+gpr+crf
This method uses stereo information.
85.56 % 74.21 % 82.81 % 88.50 % 10.12 % 11.50 % 30 s 1 core @ 2.0 Ghz (C/C++)
Z. Xiao, B. Dai, H. Li, T. Wu, X. Xu, Y. Zeng and T. Chen: Gaussian process regression-based robust free space detection for autonomous vehicle by 3-D point cloud and 2-D appearance information fusion. International Journal of Advanced Robotic Systems 2017.
69 GRES3D+SELAS
This method uses stereo information.
85.09 % 86.86 % 82.27 % 88.10 % 10.46 % 11.90 % 110 ms 4 core @ 2.8 Ghz (C/C++)
P. Shinzato: Estimation of obstacles and road area with sparse 3D points. 2015.
70 SCRFFPFHGSP
This method uses stereo information.
84.93 % 76.31 % 85.37 % 84.49 % 7.98 % 15.51 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
71 SegNet 84.04 % 78.76 % 85.50 % 82.63 % 7.72 % 17.37 % 0.01 s 8 cores @ 2.5 Ghz (C/C++)
72 Strait 84.03 % 88.71 % 82.95 % 85.13 % 9.64 % 14.87 % 69 ms GPU @ K20
73 HistonBoost
This method uses stereo information.
83.92 % 73.75 % 82.24 % 85.66 % 10.19 % 14.34 % 2.5 min >8 cores @ 3.0 Ghz (C/C++)
G. Vitor, A. Victorino and J. Ferreira: Comprehensive Performance Analysis of Road Detection Algorithms Using the Common Urban Kitti-Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application on IEEE Intelligent Vehicles Symposium (IV) 2014.
74 BM
This method uses stereo information.
83.47 % 72.23 % 75.90 % 92.72 % 16.22 % 7.28 % 2 s 2 cores @ 2.5 Ghz (Matlab)
B. WANG, V. Fremont and S. Rodriguez Florez: Color-based Road Detection and its Evaluation on the KITTI Road Benchmark. Workshop on Benchmarking Road Terrain and Lane Detection Algorithms for In-Vehicle Application, IEEE Intelligent Vehicles Symposium 2014.
75 SRF 82.44 % 87.37 % 80.60 % 84.36 % 11.18 % 15.64 % 0.2 s 1 core @ 2.5 Ghz (C/C++)
L. Xiao, B. Dai, D. Liu, D. Zhao and T. Wu: Monocular Road Detection Using Structured Random Forest. Int J Adv Robot Syst 2016.
76 SP-SS 82.36 % 72.31 % 80.48 % 84.33 % 11.27 % 15.67 % 0.01 s 4 cores @ 3.0 Ghz (Matlab + C/C++)
77 Pos-ex code 81.34 % 86.25 % 77.30 % 85.81 % 13.88 % 14.19 % 120 ms GPU(K20) @ 0.7 Ghz (Matlab + Caffe)
78 RES3D-Stereo
This method uses stereo information.
81.08 % 81.68 % 78.14 % 84.24 % 12.98 % 15.76 % 0.7 s 1 core @ 2.5 Ghz (C/C++)
P. Shinzato, D. Gomes and D. Wolf: Road estimation with sparse 3D points from stereo data. Intelligent Transportation Systems (ITSC), 2014 IEEE 17th International Conference on 2014.
79 CNN_LSTM 80.91 % 72.11 % 85.84 % 76.52 % 6.96 % 23.48 % 54 ms GPU @ 2.0 Ghz (Python + C/C++)
80 ARSL-AMI 80.36 % 70.23 % 83.24 % 77.67 % 8.61 % 22.33 % 0.05 s 4 cores @ 2.5 Ghz (C/C++)
M. Passani, J. Yebes and L. Bergasa: CRF-based semantic labeling in miniaturized road scenes . Proc. IEEE Intelligent Transportation Systems 2014.
81 SPlane + BL
This method uses stereo information.
79.63 % 83.90 % 72.59 % 88.17 % 18.34 % 11.83 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
82 CN 79.02 % 78.80 % 76.64 % 81.55 % 13.69 % 18.45 % 2 s 1 core @ 2.5 Ghz (C/C++)
J. Alvarez, T. Gevers, Y. LeCun and A. Lopez: Road Scene Segmentation from a Single Image. ECCV 2012 2012.
83 SPlane
This method uses stereo information.
78.69 % 77.16 % 71.96 % 86.80 % 18.63 % 13.20 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
84 LKW 75.53 % 79.80 % 69.68 % 82.44 % 19.75 % 17.56 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
85 ANN
This method uses stereo information.
67.70 % 52.50 % 54.19 % 90.17 % 41.98 % 9.83 % 3 s 1 core @ 3.0 Ghz (C/C++)
G. Vitor, D. Lima, A. Victorino and J. Ferreira: A 2D/3D Vision Based Approach Applied to Road Detection in Urban Environments. Intelligent Vehicles Symposium (IV), 2013 IEEE 2013.
86 VAP 62.78 % 46.54 % 48.99 % 87.41 % 50.14 % 12.59 % 1 s 1 core @ 2.5 Ghz (Matlab)
ERROR: Wrong syntax in BIBTEX file.
Table as LaTeX | Only published Methods

Lane Estimation Evaluation

UM_LANE


Method Setting Code MaxF AP PRE REC FPR FNR Runtime Environment
1 DFFA 94.31 % 88.03 % 95.33 % 93.31 % 0.80 % 6.69 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
2 AILabsLane 93.72 % 87.32 % 93.57 % 93.88 % 1.14 % 6.12 % 0.25 s GPU @ 2.5 Ghz (C/C++)
3 DCCN
This method makes use of Velodyne laser scans.
92.70 % 90.94 % 92.39 % 93.01 % 1.35 % 6.99 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
4 NVLaneNet 91.86 % 91.42 % 90.89 % 92.85 % 1.64 % 7.15 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
5 FCN-GCBs 91.24 % 85.14 % 92.15 % 90.35 % 1.35 % 9.65 % 0.08 s GPU @ 2.5 Ghz (C/C++)
6 RBNet 90.54 % 82.03 % 94.92 % 86.56 % 0.82 % 13.44 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
7 VD 90.01 % 81.60 % 88.26 % 91.82 % 2.15 % 8.18 % 4 s 1 core @ 3.0 Ghz (Matlab + C/C++)
8 Up-Conv-Poly code 89.88 % 87.52 % 92.01 % 87.84 % 1.34 % 12.16 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
9 SPRAY 83.42 % 86.84 % 84.76 % 82.12 % 2.60 % 17.88 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
10 YhY code 75.85 % 79.27 % 71.49 % 80.77 % 5.67 % 19.23 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
11 LKW 75.85 % 79.27 % 71.49 % 80.77 % 5.67 % 19.23 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
12 Strait 70.82 % 74.07 % 65.20 % 77.50 % 7.28 % 22.50 % 69 ms GPU @ K20
13 SPlane + BL
This method uses stereo information.
69.63 % 73.78 % 80.01 % 61.63 % 2.71 % 38.37 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
14 SCRFFPFHGSP
This method uses stereo information.
57.22 % 39.34 % 41.78 % 90.79 % 22.28 % 9.21 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
Table as LaTeX | Only published Methods

Behaviour Evaluation

UM_LANE


Method Setting Code PRE-20 F1-20 HR-20 PRE-30 F1-30 HR-30 PRE-40 F1-40 HR-40 Runtime Environment
1 DFFA 99.22 % 99.34 % 99.22 % 98.92 % 98.52 % 97.57 % 96.74 % 93.10 % 86.84 % 0.4 s GPU @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
2 NVLaneNet 99.22 % 99.37 % 99.01 % 99.03 % 98.62 % 97.38 % 96.74 % 93.10 % 82.89 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
3 DCCN
This method makes use of Velodyne laser scans.
99.10 % 99.08 % 98.87 % 98.79 % 98.19 % 97.38 % 96.70 % 90.70 % 89.33 % 25 ms GPU @ 2.5 Ghz (Python + C/C++)
4 AILabsLane 99.13 % 99.06 % 98.79 % 99.22 % 97.91 % 96.60 % 96.74 % 87.36 % 86.84 % 0.25 s GPU @ 2.5 Ghz (C/C++)
5 FCN-GCBs 99.18 % 99.13 % 98.76 % 99.00 % 98.11 % 96.43 % 95.56 % 93.02 % 81.58 % 0.08 s GPU @ 2.5 Ghz (C/C++)
6 RBNet 99.24 % 99.33 % 99.21 % 98.74 % 97.34 % 95.92 % 95.56 % 87.21 % 81.58 % 0.18 s GPU @ 2.5 Ghz (Matlab + C/C++)
Z. Chen and Z. Chen: RBNet: A Deep Neural Network for Unified Road and Road Boundary Detection. International Conference on Neural Information Processing 2017.
7 VD 99.12 % 98.87 % 97.71 % 99.07 % 97.93 % 95.58 % 96.70 % 90.70 % 78.67 % 4 s 1 core @ 3.0 Ghz (Matlab + C/C++)
8 Up-Conv-Poly code 99.06 % 98.84 % 98.45 % 97.57 % 95.27 % 93.14 % 90.11 % 83.72 % 77.63 % 0.08 s GPU @ 2.5 Ghz (Python + C/C++)
G. Oliveira, W. Burgard and T. Brox: Efficient Deep Methods for Monocular Road Segmentation. IROS 2016.
9 SPRAY 97.58 % 96.74 % 96.38 % 96.59 % 94.16 % 92.06 % 87.64 % 78.57 % 62.16 % 45 ms NVIDIA GTX 580 (Python + OpenCL)
T. Kuehnl, F. Kummert and J. Fritsch: Spatial Ray Features for Real-Time Ego-Lane Extraction. Proc. IEEE Intelligent Transportation Systems 2012.
10 YhY code 95.83 % 92.92 % 89.16 % 94.41 % 89.44 % 83.62 % 84.04 % 66.67 % 50.00 % 0.4 s 1 core @ 2.5 Ghz (C/C++)
ERROR: Wrong syntax in BIBTEX file.
11 LKW 95.83 % 92.92 % 89.16 % 94.41 % 89.44 % 83.62 % 84.04 % 66.67 % 50.00 % 0.1 s 1 core @ 2.5 Ghz (C/C++)
12 SPlane + BL
This method uses stereo information.
95.53 % 92.88 % 91.21 % 91.89 % 87.12 % 74.28 % 79.79 % 47.13 % 0.00 % 2 s 1 core @ 3.0 Ghz (C/C++)
N. Einecke and J. Eggert: Block-Matching Stereo with Relaxed Fronto-Parallel Assumption. IV 2014.
13 Strait 94.97 % 91.74 % 88.38 % 89.60 % 81.85 % 73.89 % 67.02 % 48.28 % 32.89 % 69 ms GPU @ K20
14 SCRFFPFHGSP
This method uses stereo information.
94.88 % 87.95 % 82.98 % 87.91 % 78.90 % 71.95 % 60.64 % 43.68 % 38.16 % 5 s 8 cores @ 2.5 Ghz (C/C++, Matlab)
I. Gheorghe: Semantic Segmentation of Terrain and Road Terrain for Advanced Driver Assistance Systems. 2015.
Table as LaTeX | Only published Methods

Related Datasets

Citation

When using this dataset in your research, we will be happy if you cite us:
@INPROCEEDINGS{Fritsch2013ITSC,
  author = {Jannik Fritsch and Tobias Kuehnl and Andreas Geiger},
  title = {A New Performance Measure and Evaluation Benchmark for Road Detection Algorithms},
  booktitle = {International Conference on Intelligent Transportation Systems (ITSC)},
  year = {2013}
}



eXTReMe Tracker