Are we ready for autonomous driving? The KITTI vision benchmark suite

By Andreas Geiger, Philip Lenz, Raquel Urtasun. 2012

15,381 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Learning Transferable Visual Models From Natural Language Supervision (2021, background): “, 2011), the KITTI dataset (Geiger et al., 2012), PatchCamelyon (Veeling et al.”
Semantic Understanding of Scenes Through the ADE20K Dataset (2016, background): “t classiﬁcation/detection datasets. Most of the large-scale datasets typically only contain labels at the image level or provide bounding boxes. Examples include ImageNet [29], Pascal [10], and KITTI [11]. ImageNet has the largest set of classes, but contains relatively simple scenes. Pascal and KIT”
Cascade R-CNN: High Quality Object Detection and Instance Segmentation (2019, methodology): “Finally,we further present the results of a larger set of experiments, performed on various popular generic/specific object detection datasets, including PASCAL VOC [12], KITTI [15], CityPerson [63] and WiderFace [60].”
Objects as Points (2019, methodology): “3D bounding box estimation powers autonomous driving [17].”
DeepSetNet: Predicting Sets with Deep Neural Networks (2016, methodology): “We use the multi-scale deep CNN approach (MS-CNN) [3] trained on the KITTI dataset [13] for our Table 2: Count mean absolute error on UCSD dataset.”
COCO-Stuff: Thing and Stuff Classes in Context (2016, background): “In particular, MSRC 21 [46], KITTI [23], CamVid [7], SIFT Flow [36] and NYUD [47] all have less than 5,000 images (Table 1).”
EPIC-KITCHENS VISOR Benchmark: VIdeo Segmentations and Object Relations (2022, background): “Rich pixel-level annotations have transformed image understanding [17, 23] and autonomous driving [10, 13], amongst other tasks.”
Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100 (2020, background): “with multiple annotations and tasks, towards learning intermediate representations through downstream and multi-task learning on the same input. This has been recently achieved for autonomous driving [17,18,19,20,21,22,23,24] and scene understanding [25,26]. For example, Taskonomy [25] contains 26 t”
WoodScape: A Multi-Task, Multi-Camera Fisheye Dataset for Autonomous Driving (2019, methodology): “Performance of 2D object detection is reported in terms of mean average precision (mAP) when IoU≥ 0.5 between predicted and ground truth bounding boxes.”
Scene Parsing through ADE20K Dataset (2017, background): “Examples include Imagenet [26], Pascal [10], and KITTI [12].”
Learning Image Representations Tied to Ego-Motion (2015, methodology): “To this end, we cast the problem in terms of unsupervised equivariant feature learning.”
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation (2015, methodology): “interesting result here is the large performance improvement in class average and mIOU metrics that is obtained when a large training dataset, obtained by combining [21], [25], [59], [60], is used to train SegNet.”
VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (2017, methodology): “Our experimental setup is based on the LiDAR specifications of the KITTI dataset [11].”
Robust Physical-World Attacks on Deep Learning Models (2017, background): “on many computer vision tasks [11, 14, 36]. Based on These authors contributed equally. these successes, they are increasingly being used as part of control pipelines in physical systems such as cars [8,17], UAVs [4,24], and robots [40]. Recent work, however, has demonstrated that DNNs are vulnerabl”
IRS: A Large Synthetic Indoor Robotics Stereo Dataset for Disparity and Surface Normal Estimation (2019, background): “Existing studies [10], [11], [12], [13] proposed stereo datasets collected by real sensing hardware, which contributed a lot to stereo vision research.”
Benchmark and Survey of Automated Machine Learning Frameworks (2019, methodology): “Using agreed-on data sets with standardized search spaces for benchmarking, like it is done in other fields of research (e.g., Geiger et al., 2012), would increase the comparability.”