# How To Train Your Deep Multi-Object Tracker

arXiv: https://arxiv.org/abs/1906.06618 · HTML: https://mlsanity.com/paper/1906.06618/

## Key facts

- Cited by 71 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 15 Jun 2019.

## Built on by (most cited first, 30 of 71)

- [BoT-SORT: Robust Associations Multi-Pedestrian Tracking](https://mlsanity.com/paper/2206.14651/) (2022)
- [StrongSORT: Make DeepSORT Great Again](https://mlsanity.com/paper/2202.13514/) (2022)
- [SimpleTrack: Understanding and Rethinking 3D Multi-object Tracking](https://mlsanity.com/paper/2111.09621/) (2021)
- [TransMOT: Spatial-Temporal Graph Transformer for Multiple Object Tracking](https://mlsanity.com/paper/2104.00194/) (2021)
- [Towards Grand Unification of Object Tracking](https://mlsanity.com/paper/2207.07078/) (2022)
- [EagerMOT: 3D Multi-Object Tracking via Sensor Fusion](https://mlsanity.com/paper/2104.14682/) (2021)
- [MOTSynth: How Can Synthetic Data Help Pedestrian Detection and Tracking?](https://mlsanity.com/paper/2108.09518/) (2021)
- [4D Panoptic LiDAR Segmentation](https://mlsanity.com/paper/2102.12472/) (2021)
- [SiamMOT: Siamese Multi-Object Tracking](https://mlsanity.com/paper/2105.11595/) (2021)
- [Unifying Short and Long-Term Tracking with Graph Hierarchies](https://mlsanity.com/paper/2212.03038/) (2022)
- [DEFT: Detection Embeddings for Tracking](https://mlsanity.com/paper/2102.02267/) (2021)
- [RelationTrack: Relation-aware Multiple Object Tracking with Decoupled Representation](https://mlsanity.com/paper/2105.04322/) (2021)
- [Learnable Graph Matching: Incorporating Graph Partitioning with Deep Feature Learning for Multiple Object Tracking](https://mlsanity.com/paper/2103.16178/) (2021)
- [Visual Object Tracking with Discriminative Filters and Siamese Networks: A Survey and Outlook](https://mlsanity.com/paper/2112.02838/) (2021)
- [GIAOTracker: A comprehensive framework for MCMOT with global information and optimizing strategies in VisDrone 2021](https://mlsanity.com/paper/2202.11983/) (2022)
- [Simple Cues Lead to a Strong Multi-Object Tracker](https://mlsanity.com/paper/2206.04656/) (2022)
- [Learning a Proposal Classifier for Multiple Object Tracking](https://mlsanity.com/paper/2103.07889/) (2021)
- [TransCenter: Transformers with Dense Representations for Multiple-Object Tracking](https://mlsanity.com/paper/2103.15145/) (2021)
- [AnimalTrack: A Benchmark for Multi-Animal Tracking in the Wild](https://mlsanity.com/paper/2205.00158/) (2022)
- [LMGP: Lifted Multicut Meets Geometry Projections for Multi-Camera Multi-Object Tracking](https://mlsanity.com/paper/2111.11892/) (2021)
- [DyGLIP: A Dynamic Graph Model with Link Prediction for Accurate Multi-Camera Multiple Object Tracking](https://mlsanity.com/paper/2106.06856/) (2021)
- [PolarMOT: How Far Can Geometric Relations Take Us in 3D Multi-Object Tracking?](https://mlsanity.com/paper/2208.01957/) (2022)
- [Quo Vadis: Is Trajectory Forecasting the Key Towards Long-Term Multi-Object Tracking?](https://mlsanity.com/paper/2210.07681/) (2022)
- [Track without Appearance: Learn Box and Tracklet Embedding with Local and Global Motion Patterns for Vehicle Tracking](https://mlsanity.com/paper/2108.06029/) (2021)
- [ReST: A Reconfigurable Spatial-Temporal Graph Model for Multi-Camera Multi-Object Tracking](https://mlsanity.com/paper/2308.13229/) (2023)
- [Discriminative Appearance Modeling with Multi-track Pooling for Real-time Multi-object Tracking](https://mlsanity.com/paper/2101.12159/) (2021)
- [Tracking People with 3D Representations](https://mlsanity.com/paper/2111.07868/) (2021)
- [Looking Beyond Two Frames: End-to-End Multi-Object Tracking Using Spatial and Temporal Transformers](https://mlsanity.com/paper/2103.14829/) (2021)
- [A Comprehensive Review of Computer Vision in Sports: Open Issues, Future Trends and Research Directions](https://mlsanity.com/paper/2203.02281/) (2022)
- [Tracking People by Predicting 3D Appearance, Location & Pose](https://mlsanity.com/paper/2112.04477/) (2021)

## Abstract (arXiv)

The recent trend in vision-based multi-object tracking (MOT) is heading towards leveraging the representational power of deep learning to jointly learn to detect and track objects. However, existing methods train only certain sub-modules using loss functions that often do not correlate with established tracking evaluation measures such as Multi-Object Tracking Accuracy (MOTA) and Precision (MOTP). As these measures are not differentiable, the choice of appropriate loss functions for end-to-end training of multi-object tracking methods is still an open research problem. In this paper, we bridge this gap by proposing a differentiable proxy of MOTA and MOTP, which we combine in a loss function suitable for end-to-end training of deep multi-object trackers. As a key ingredient, we propose a Deep Hungarian Net (DHN) module that approximates the Hungarian matching algorithm. DHN allows estimating the correspondence between object tracks and ground truth objects to compute differentiable proxies of MOTA and MOTP, which are in turn used to optimize deep trackers directly. We experimentally demonstrate that the proposed differentiable framework improves the performance of existing multi-object trackers, and we establish a new state of the art on the MOTChallenge benchmark. Our code is publicly available from https://github.com/yihongXU/deepMOT.