README.md

metadata

license: mit
tags:
  - object-detection
  - object-tracking
  - video
  - video-object-segmentation
inference: false

unicorn_track_tiny_mask

Table of Contents

• unicorn_track_tiny_mask
  • Table of Contents
  • Model Details
  • Uses
    • Direct Use
  • Evaluation Results

Model Details

Unicorn accomplishes the great unification of the network architecture and the learning paradigm for four tracking tasks. Unicorn puts forwards new state-of-the-art performance on many challenging tracking benchmarks using the same model parameters. This model has an input size of 800x1280.

• License: This model is licensed under the MIT license
• Resources for more information:
  • Research Paper
  • GitHub Repo

Uses

Direct Use

This model can be used for:

• Single Object Tracking (SOT)
• Multiple Object Tracking (MOT)
• Video Object Segmentation (VOS)
• Multi-Object Tracking and Segmentation (MOTS)

Evaluation Results

LaSOT AUC (%): 67.7 BDD100K mMOTA (%): 39.9 DAVIS17 J&F (%): 68.0 BDD100K MOTS mMOTSA (%): 29.7

Citation Information

@inproceedings{unicorn,
  title={Towards Grand Unification of Object Tracking},
  author={Yan, Bin and Jiang, Yi and Sun, Peize and Wang, Dong and Yuan, Zehuan and Luo, Ping and Lu, Huchuan},
  booktitle={ECCV},
  year={2022}
}