Awesome video understanding toolkits based on PaddlePaddle. It supports video data annotation tools, lightweight RGB and skeleton based action recognition model, practical applications for video tagging and sport action detection.

Code release for ActionFormer (ECCV 2022)

[ICLR 2022] TAda! Temporally-Adaptive Convolutions for Video Understanding. This codebase provides solutions for video classification, video representation learning and temporal detection.

Official Tensorflow Implementation of the AAAI-2020 paper "Temporally Grounding Language Queries in Videos by Contextual Boundary-aware Prediction"

Code for Spotting Temporally Precise, Fine-Grained Events in Video

Weakly-supervised Action Localization

Adversarial Background-Aware Loss for Weakly-supervised Temporal Activity Localization (ECCV 2020)

Code for Look for the Change paper published at CVPR 2022

Action-Localization, Atomic Visual Actions (AVA) Dataset

[BMVC 2021]: Official PyTorch implementation of : "Few Shot Temporal Action Localization using Query Adaptive Transformers"

Tensorflow Reproduction of the EMNLP-2018 paper "Temporally Grounding Natural Sentence in Video"

A PyTorch implementation of ACRNet based on ICME 2023 paper "Weakly-supervised Temporal Action Localization with Adaptive Clustering and Refining Network"

Official repository of ECCV 2024 paper - "HAT: History-Augmented Anchor Transformer for Online Temporal Action Localization"

Code for the paper "Multi-Task Learning of Object States and State-Modifying Actions from Web Videos" published in TPAMI

[ICCV 2021] Official PyTorch implementation for "D2-Net: Weakly-Supervised Action Localization via Discriminative Embeddings and Denoised Activations"

Localizing narrated human activities in lifestyle vlogs.

Video Query System