fortunto2/Awesome-Deep-Video-Research

Awesome Deep Video Research

• Awesome Deep Vision: https://github.com/kjw0612/awesome-deep-vision

DATASETS

VIDEO SUMMARIZATION

• TVSum dataset: https://github.com/yalesong/tvsum
• SumMe dataset: https://people.ee.ethz.ch/~gyglim/vsum/#benchmark
• OVP and YouTube datasets: https://sites.google.com/site/vsummsite/

HIGHLIGHTS

• BAIDU: http://ai.baidu.com/broad/introduction

Action Recognition

• UCF101: http://crcv.ucf.edu/data/UCF101.php

TOOLS

• Vireo is a lightweight and versatile video processing library written in C++11: https://github.com/twitter/vireo
• Scanner: Efficient Video Analysis at Scale: https://github.com/scanner-research/scanner
• Esper from Scannser: https://github.com/scanner-research/esper
• Kitware Image and Video Exploitation and Retrieval: https://github.com/Kitware/kwiver
• A distributed visual search and visual data analytics platform: https://github.com/AKSHAYUBHAT/DeepVideoAnalytics
• VGG Image Search Engine (VISE): http://www.robots.ox.ac.uk/~vgg/software/vise/

API

• AWS Video Frame-based Analysis: https://github.com/awslabs/video-frame-based-analysis
• Microsoft https://github.com/Microsoft/Cognitive-Samples-VideoFrameAnalysis

Dataset Tagging GUI

• Visual Object Tagging Tool: An electron app for building end to end Object Detection Models from Images and Videos. https://github.com/Microsoft/VoTT

• Universal Data Tool: Collaborate & label video and images in an easy web interface or desktop app. https://github.com/UniversalDataTool/universal-data-tool