/TAP

Temporal Alignment Prediction for Supervised Representation Learning and Few-Shot Sequence Classification

Primary LanguagePython

Temporal Alignment Prediction for Supervised Representation Learning and Few-Shot Sequence Classification

  1. Introduction.

This package includes the python codes for implementing Temporal Alignment Prediction (TAP) in supervised learning and few-shot action recognition, described in

Bing Su and Ji-Rong Wen. "Temporal Alignment Prediction for Supervised Representation Learning and Few-Shot Sequence Classification" ICLR, 2022.

  1. License & disclaimer.

    The codes can be used for research purposes only. This package is strictly for non-commercial academic use only.

  2. Usage

    The "Supervised_Learning" folder contains codes for using TAP in supervised distance learning for sequence data.

    The "Few_Shot_Learning" folder contains codes for using TAP in few-shot action recognition.

    Please refer to the ReadMe files in the folds.

  3. Citations

Please cite the following paper if you use the codes:

Bing Su and Ji-Rong Wen, "Temporal Alignment Prediction for Supervised Representation Learning and Few-Shot Sequence Classification", ICLR, 2022.