Semantic Segmentation Evaluation

This is the KITTI semantic segmentation benchmark. It consists of 200 semantically annotated train as well as 200 test images corresponding to the KITTI Stereo and Flow Benchmark 2015. The data format and metrics are conform with The Cityscapes Dataset.

The data can be downloaded here:

Note: On 12.04.2018 we have fixed several annotation errors in the dataset, please download the dataset again if you have an old version.

Our evaluation table ranks all methods according to the PASCAL VOC intersection-over-union metric (IoU). IoU = TP/(TP+FP+FN), where TP, FP, and FN are the numbers of true positive, false positive, and false negative pixels, respectively. Like in cityscapes we also use an instance-level intersection over union iIoU = iTP/(iTP+FP+iFN). In contrast to the standard IoU measure, iTP and iFN are computed by weighting the contribution of each pixel by the ratio of the class’ average instance size to the size of the respective ground truth instance.

  • IoU class:  Intersection over Union for each class IoU=TP/(TP+FP+FN)
  • iIoU class:    Instance Intersection over Union iIoU=iTP/(iTP+FP+iFN)
  • IoU category:   Intersection over Union for each category IoU=TP/(TP+FP+FN)
  • iIoU category:     Instance Intersection over Union for each category iIoU=iTP/(iTP+FP+iFN)

Important Policy Update: As more and more non-published work and re-implementations of existing work is submitted to KITTI, we have established a new policy: from now on, only submissions with significant novelty that are leading to a peer-reviewed paper in a conference or journal are allowed. Minor modifications of existing algorithms or student research projects are not allowed. Such work must be evaluated on a split of the training set. To ensure that our policy is adopted, new users must detail their status, describe their work and specify the targeted venue during registration. Furthermore, we will regularly delete all entries that are 6 months old but are still anonymous or do not have a paper associated with them. For conferences, 6 month is enough to determine if a paper has been accepted and to add the bibliography information. For longer review cycles, you need to resubmit your results.
Additional information used by the methods
  • Laser Points: Method uses point clouds from Velodyne laser scanner
  • Depth: Method uses depth from stereo.
  • Video: Method uses 2 or more temporally adjacent images
  • Additional training data: Use of additional data sources for training (see details)

Method Setting Code IoU class iIoU class IoU category iIoU category Runtime Environment
1 WRP code 76.44 50.92 89.63 73.69 1 s GPU @ >3.5 Ghz (Python)
A. Ganeshan, A. Vallet, Y. Kudo, S. Maeda, T. Kerola, R. Ambrus, D. Park and A. Gaidon: Warp-Refine Propagation: Semi-Supervised Auto-labeling via Cycle- consistency. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021.
2 UJS_model 75.11 47.71 89.53 75.75 0.26 s 1 core @ 2.5 Ghz (Python + C/C++)
L. Yingfeng Cai and Z. Hai Wang: Multi-Target Pan-Class Intrinsic Relevance Driven Model for Improving Semantic Segmentation in Autonomous Driving. IEEE Transactions on Image Processing (TIP) 2021.
3 RoadFormer+ 73.13 45.88 88.75 73.46 0.04 s 1 core @ 2.5 Ghz (C/C++)
4 VideoProp-LabelRelax 72.82 48.68 88.99 75.26 n s GPU @ 1.5 Ghz (Python + C/C++)
B. Yi Zhu*: Improving Semantic Segmentation via Video Propagation and Label Relaxation. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) 2019.
5 SN_DN161_fat_pyrx8 68.89 40.45 87.06 67.93 1 s 6 x Tesla V100
P. Bevandić, M. Oršić, I. Grubišić, J. Šarić and S. Šegvić: Multi-domain semantic segmentation with overlapping labels. WACV 2022.
6 MSeg1080_RVC code 62.64 31.62 86.59 68.05 0.49 s 1 core @ 3.0 Ghz (Python)
J. Lambert, Z. Liu, O. Sener, J. Hays and V. Koltun: MSeg: A Composite Dataset for Multi- domain Semantic Segmentation. Computer Vision and Pattern Recognition (CVPR) 2020.
7 Chroma UDA 60.36 31.70 80.73 61.91 0.4 s GPU @ 2.5 Ghz (Python)
O. Erkent and C. Laugier: Semantic Segmentation with Unsupervised Domain Adaptation Under Varying Weather Conditions for Autonomous Vehicles. IEEE Robotics and Automation Letters 2020.
8 IfN-DomAdap-Seg 59.50 30.28 81.57 61.91 1 s GPU @ 2.0 Ghz (Python)
J. Bolte, M. Kamp, A. Breuer, S. Homoceanu, P. Schlicht, F. Hüger, D. Lipinski and T. Fingscheidt: Unsupervised Domain Adaptation to Improve Image Segmentation Quality Both in the Source and Target Domain. Proc. of CVPR - Workshops 2019.
9 SegStereo code 59.10 28.00 81.31 60.26 0.6 s Nvidia GTX Titan Xp
G. Yang, H. Zhao, J. Shi, Z. Deng and J. Jia: SegStereo: Exploiting Semantic Information for Disparity Estimation. ECCV 2018.
10 MCANet code 58.52 24.00 83.04 54.06 0.003 s 1 GPU NVIDIA geForce RTX
T. Singha, D. Pham and A. Krishna: Multi-encoder Context Aggregation Network for Structured and Unstructured Urban Street Scene Analysis. IEEE Access 2023.
11 SDBNetV2 56.77 23.11 81.08 50.77 0.004 s 1 GPU NVIDIA geForce RTX
T. Singha, D. Pham and A. Krishna: Improved Short-term Dense Bottleneck network for efficient scene analysis. Computer Vision and Image Understanding 2023.
12 EPNet 54.59 24.24 79.47 51.83 0.03 s NVIDIA 4090 (Python)
13 SGDepth code 53.04 24.36 78.65 55.95 0.1 s GPU @ 2.5 Ghz (Python)
M. Klingner, J. Termöhlen, J. Mikolajczyk and T. Fingscheidt: Self-Supervised Monocular Depth Estimation: Solving the Dynamic Object Problem by Semantic Guidance. ECCV 2020.
14 SDBNet code 51.80 18.72 78.00 44.46 0.01 s 1 GPU NVIDIA geForce RTX
T. Singha, D. Pham and A. Krishna: SDBNet: Lightweight Real-Time Semantic Segmentation Using Short-Term Dense Bottleneck. Proc. DICTA 2022.
15 SDNet code 51.14 17.74 79.62 50.45 0.2 s GPU @ 2.5 Ghz (Python + C/C++)
M. Ochs, A. Kretz and R. Mester: SDNet: Semantic Guided Depth Estimation Network. German Conference on Pattern Recognition (GCPR) 2019.
16 test 50.07 20.41 73.86 40.32 1 s 1 core @ 2.5 Ghz (C/C++)
17 SFRSeg code 49.27 17.43 77.91 46.88 0.005 s 1 GPU NVIDIA geForce RTX
T. Singha, D. Pham and A. Krishna: A real-time semantic segmentation model using iteratively shared features in multiple sub- encoders. Pattern Recognition 2023.
18 APMoE_seg_ROB code 47.96 17.86 78.11 49.17 0.2 s GPU @ 3.5 Ghz (Matlab/C++)
S. Kong and C. Fowlkes: Pixel-wise Attentional Gating for Parsimonious Pixel Labeling. arxiv 1805.01556 2018.
19 LIISIESS 46.73 19.82 76.04 49.91 NA s 1 core @ 2.5 Ghz (C/C++)
L. Sun, J. Bockman and S. Changming: A Framework for Leveraging Inter-image Information in Stereo Images for Enhanced Semantic Segmentation. IEEE Transactions on Instrumentation and Measurement 2023.
Table as LaTeX | Only published Methods

Related Datasets

  • The Cityscapes Dataset: The cityscapes dataset was recorded in 50 German cities and offers high quality pixel-level annotations of 5 000 frames in addition to a larger set of 20 000 weakly annotated frames.
  • Wilddash: Wilddash is a benchmark for semantic and instance segmentation. It aims to improve the expressiveness of performance evaluation for computer vision algorithms in regard to their robustness under real-world conditions.


When using this dataset in your research, we will be happy if you cite us:
  author = { AlhaijaandHassan and MustikovelaandSiva and MeschederandLars and GeigerandAndreas and RotherandCarsten},
  title = {Augmented Reality Meets Computer Vision: Efficient Data Generation for Urban Driving Scenes},
  journal = {International Journal of Computer Vision (IJCV)},
  year = {2018}

eXTReMe Tracker