eriklindernoren / Action Recognition
Exploration of different solutions to action recognition in video, using neural networks implemented in PyTorch.
Stars: ✭ 129
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Action Recognition
Epic Kitchens 55 Action Models
EPIC-KITCHENS-55 baselines for Action Recognition
Stars: ✭ 68 (-47.29%)
Mutual labels: action-recognition
Temporal Segment Networks
Code & Models for Temporal Segment Networks (TSN) in ECCV 2016
Stars: ✭ 1,287 (+897.67%)
Mutual labels: action-recognition
Keras Kinetics I3d
keras implementation of inflated 3d from Quo Vardis paper + weights
Stars: ✭ 116 (-10.08%)
Mutual labels: action-recognition
Tdn
[CVPR 2021] TDN: Temporal Difference Networks for Efficient Action Recognition
Stars: ✭ 72 (-44.19%)
Mutual labels: action-recognition
M Pact
A one stop shop for all of your activity recognition needs.
Stars: ✭ 85 (-34.11%)
Mutual labels: action-recognition
Training toolbox caffe
Training Toolbox for Caffe
Stars: ✭ 51 (-60.47%)
Mutual labels: action-recognition
I3d finetune
TensorFlow code for finetuning I3D model on UCF101.
Stars: ✭ 128 (-0.78%)
Mutual labels: action-recognition
Video classification pytorch
Video Classification based on PyTorch
Stars: ✭ 89 (-31.01%)
Mutual labels: action-recognition
Modelfeast
Pytorch model zoo for human, include all kinds of 2D CNN, 3D CNN, and CRNN
Stars: ✭ 116 (-10.08%)
Mutual labels: action-recognition
Daps
This repo allocate DAPs code of our ECCV 2016 publication
Stars: ✭ 74 (-42.64%)
Mutual labels: action-recognition
Vidvrd Helper
To keep updates with VRU Grand Challenge, please use https://github.com/NExTplusplus/VidVRD-helper
Stars: ✭ 81 (-37.21%)
Mutual labels: action-recognition
Tdd
Trajectory-pooled Deep-Convolutional Descriptors
Stars: ✭ 99 (-23.26%)
Mutual labels: action-recognition
Hake Action
As a part of the HAKE project, includes the reproduced SOTA models and the corresponding HAKE-enhanced versions (CVPR2020).
Stars: ✭ 72 (-44.19%)
Mutual labels: action-recognition
Epic Kitchens 55 Annotations
🍴 Annotations for the EPIC KITCHENS-55 Dataset.
Stars: ✭ 120 (-6.98%)
Mutual labels: action-recognition
Fight detection
Real time Fight Detection Based on 2D Pose Estimation and RNN Action Recognition
Stars: ✭ 65 (-49.61%)
Mutual labels: action-recognition
Video Dataset Loading Pytorch
Generic PyTorch Dataset Implementation for Loading, Preprocessing and Augmenting Video Datasets
Stars: ✭ 92 (-28.68%)
Mutual labels: action-recognition
Mmaction
An open-source toolbox for action understanding based on PyTorch
Stars: ✭ 1,711 (+1226.36%)
Mutual labels: action-recognition
Skeleton Based Action Recognition Papers And Notes
Skeleton-based Action Recognition Papers and Small Notes and Top 2 Leaderboard for NTU-RGBD
Stars: ✭ 126 (-2.33%)
Mutual labels: action-recognition
Movienet Tools
Tools for movie and video research
Stars: ✭ 113 (-12.4%)
Mutual labels: action-recognition
Action Recognition in Video
This repo will serve as a playground where I investigate different approaches to solving the problem of action recognition in video.
I will mainly use the UCF-101 dataset.
Setup
$ cd data/
$ bash download_ucf101.sh # Downloads the UCF-101 dataset (~7.2 GB)
$ unrar x UCF101.rar # Unrars dataset
$ unzip ucfTrainTestlist.zip # Unzip train / test split
$ python3 extract_frames.py # Extracts frames from the video (~26.2 GB, go grab a coffee for this)
ConvLSTM
The only approach investigated so far. Enables action recognition in video by a bi-directional LSTM operating on frame embeddings extracted by a pre-trained ResNet-152 (ImageNet).
The model is composed of:
- A convolutional feature extractor (ResNet-152) which provides a latent representation of video frames
- A bi-directional LSTM classifier which based on the latent representation of the video predicts the activity depicted
I have made a trained model available here.
Train
$ python3 train.py --dataset_path data/UCF-101-frames/ \
--split_path data/ucfTrainTestlist \
--num_epochs 200 \
--sequence_length 40 \
--img_dim 112 \
--latent_dim 512
Test on Video
$ python3 test_on_video.py --video_path data/UCF-101/SoccerPenalty/v_SoccerPenalty_g01_c01.avi \
--checkpoint_model model_checkpoints/ConvLSTM_150.pth
Results
The model reaches a classification accuracy of 91.27% accuracy on a randomly sampled test set, composed of 20% of the total amount of video sequences from UCF-101. Will re-train this model on the offical train / test splits and post results as soon as I have time.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].