harshitrajora / Video-Captioning Public

Notifications You must be signed in to change notification settings
Fork 0
Star 1

This project is an End-to-End Video Captioning System designed to bridge the gap between Computer Vision and Natural Language Processing. It automatically generates descriptive text for video content, essentially teaching a computer to "watch" a video and describe what is happening in English

1 star 0 forks Branches Tags Activity

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
Desktop/video_captioningproject		Desktop/video_captioningproject
.DS_Store		.DS_Store
README.md		README.md

Repository files navigation

Video_Captioning

Implemented Syntax Aware LSTM Model capable of generating captions on video clips using Computer Vision And NLP.

Preprocessed the MSVD dataset and used pretrained Inception V4 model as encoder generating features for each video frames.

Trained LSTM with Attention from scratch in Pytorch and used it as Decoder obtaining overall BLEU score of 34%

About

This project is an End-to-End Video Captioning System designed to bridge the gap between Computer Vision and Natural Language Processing. It automatically generates descriptive text for video content, essentially teaching a computer to "watch" a video and describe what is happening in English

natural-language-processing computer-vision deep-learning data-eng

Report repository

Releases

No releases published

Packages

No packages published

Languages

Python 100.0%