MOT20: A benchmark for multi object tracking in crowded scenes

By Patrick Dendorfer, Hamid Rezatofighi, Anton Milan, Javen Qinfeng Shi, Daniel Cremers, Ian David Reid, Stefan Roth, Konrad Schindler, L. Leal-Taix'e. arXiv.org, 2020

875 citations recorded by the source corpus.

Standardized benchmarks are crucial for the majority of computer vision applications. Although leaderboards and ranking tables should not be over-claimed, benchmarks often provide the most objective measure of performance and are therefore important guides for research. The benchmark for Multiple Object Tracking, MOTChallenge, was launched with the goal to establish a standardized evaluation of multiple object tracking methods. The challenge focuses on multiple people tracking, since pedestrians are well studied in the tracking community, and precise tracking and detection has high practical relevance. Since the first release, MOT15, MOT16, and MOT17 have tremendously contributed to the community by introducing a clean dataset and precise framework to benchmark multi-object trackers. In this paper, we present our MOT20benchmark, consisting of 8 new sequences depicting very crowded challenging scenes. The benchmark was presented first at the 4thBMTT MOT Challenge Workshop at the Computer Vision and Pattern Recognition Conference (CVPR) 2019, and gives to chance to evaluate state-of-the-art methods for multiple object tracking when handling extremely crowded scenarios.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/d11c50e49998e2156da7c179a3caea86e9601abd

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Global and Planetary Change

Citations in the corpus

TrackFormer: Multi-Object Tracking with Transformers (2021, methodology): “Comparison of multi-object tracking methods on the MOT17 [30] and MOT20 [13] test sets.”
Rethinking the Competition Between Detection and ReID in Multiobject Tracking (2020, methodology): “We present the evaluation results with the comparisons to recently-prevailing one-shot trackers on MOT16 [33], MOT17 [33] and MOT20 [38].”
StrongSORT: Make DeepSORT Great Again (2022, methodology): “We conduct experiments on the MOT17 [31] and MOT20 [9] datasets under the “private detection” protocol.”
BoT-SORT: Robust Associations Multi-Pedestrian Tracking (2022, methodology): “Experiments were conducted on two of the most popular benchmarks in the ﬁeld of multi-object tracking for pedestrian detection and tracking in unconstrained environments: MOT17 [29] and MOT20 [11] datasets under the “private detection” protocol.”
HOTA: A Higher Order Metric for Evaluating Multi-object Tracking (2020, methodology): “MOTA was adopted for evaluation in the PETS workshop series [23] and remains, to this day, the most commonly used metric for evaluating MOT algorithms, although it has often been highly criticised [65,6,43,40,55,60,18,48,51,76, 52,88,20,50] for its bias toward overemphasizing detection over associat”
Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking (2022, methodology): “MOT17 [43] and MOT20 [14] are for pedestrian tracking, whose motions are nearly linear.”
DanceTrack: Multi-Object Tracking in Uniform Appearance and Diverse Motion (2021, background): “The first property makes tracking by purely comparing object appearance invalid because the exDataset MOT17 [27] MOT20 [11] DanceTrack”
Deep OC-Sort: Multi-Pedestrian Tracking by Adaptive Re-Identification (2023, methodology): “We conduct experiments on MOT17 [5], MOT20 [6], and DanceTrack [7].”
Joint Object Detection and Multi-Object Tracking with Graph Neural Networks (2021, methodology): “We evaluate our method on MOTChallenges [64]–[67], including the 2DMOT15/MOT16/MOT17/MOT20 benchmarks.”
How many Observations are Enough? Knowledge Distillation for Trajectory Forecasting (2022, neutral): “In real-time applications, a visual tracking system may provide inaccurate observations for sequences of such length [11]: occlusions, false detections and non-rigid shape deformations pose non-trivial issues.”