Timezone: »
In this paper, we propose a self-supervised learning procedure for training a robust multi-object tracking (MOT) model given only unlabeled video. While several self-supervisory learning signals have been proposed in prior work on single-object tracking, such as color propagation and cycle-consistency, these signals are not effective for training RNN models, which are needed to achieve accurate MOT: they yield degenerate models that, for instance, always match new detections to tracks with the closest initial detections. We propose a novel self-supervisory signal that we call cross-input consistency: we construct two distinct inputs for the same sequence of video, by hiding different information about the sequence in each input. We then compute tracks in that sequence by applying an RNN model independently on each input, and train the model to produce consistent tracks across the two inputs. We evaluate our unsupervised method on MOT17 and KITTI --- remarkably, we find that, despite training only on unlabeled video, our unsupervised approach outperforms four supervised methods published in the last 1--2 years, including Tracktor++, FAMNet, GSM, and mmMOT.
Author Information
Favyen Bastani (Massachusetts Institute of Technology)
Songtao He (Massachusetts Institute of Technology)
Samuel Madden (Massachusetts Institute of Technology)
More from the Same Authors
-
2021 : Synthesizing Video Trajectory Queries »
Stephen Mell · Favyen Bastani · Stephan Zdancewic · Osbert Bastani -
2022 : Generic Entity Resolution Models »
Jiawei Tang · Yifei Zuo · Lei Cao · Samuel Madden -
2022 : RoTaR: Efficient Row-Based Table Representation Learning via Teacher-Student Training (Short Paper) »
Zui Chen · Lei Cao · Samuel Madden -
2022 : Invited Talk - Favyen Bastani - Allen Institute for AI »
Favyen Bastani -
2019 Poster: Park: An Open Platform for Learning-Augmented Computer Systems »
Hongzi Mao · Parimarjan Negi · Akshay Narayan · Hanrui Wang · Jiacheng Yang · Haonan Wang · Ryan Marcus · Ravichandra Addanki · Mehrdad Khani Shirkoohi · Songtao He · Vikram Nathan · Frank Cangialosi · Shaileshh Venkatakrishnan · Wei-Hung Weng · Song Han · Tim Kraska · Dr.Mohammad Alizadeh