abhshkdz/neural-vqa
A Torch implementation of a neural visual question answering model combining CNN image features with LSTM language processing.
Not currently ranked — collecting fresh signals.
star history
This repository implements the VIS+LSTM visual question answering model from a research paper by Ren, Kiros & Zemel. It uses VGG-19 CNN to extract image features and LSTM to process questions, combining both to generate answers about images. The model is trained on MSCOCO images paired with VQA dataset question-answer pairs.
Frequently asked
- What is abhshkdz/neural-vqa?
- A Torch implementation of a neural visual question answering model combining CNN image features with LSTM language processing.
- Is neural-vqa open source?
- Yes — abhshkdz/neural-vqa is an open-source project tracked on heatdrop.
- What language is neural-vqa written in?
- abhshkdz/neural-vqa is primarily written in Lua.
- How popular is neural-vqa?
- abhshkdz/neural-vqa has 486 stars on GitHub.
- Where can I find neural-vqa?
- abhshkdz/neural-vqa is on GitHub at https://github.com/abhshkdz/neural-vqa.