sumanthd17 / Image-caption-generator

Image caption generator with Encoder-Decoder Network. Create a flask API for uploading images and generating the captions

Image Caption Generator

The goal of the project is to generate captions for images by learning visual representations from Images

Implementation:

MS COCO dataset has 5 captions for each image in the dataset. These are rich dense captions which contains information about the image (descriptive and positional). This project used a ResNet50 backbone for obtaining the visual features and LSTM Decoder for generating the captions.

Dataset

MS COCO - Image Captioning Dataset

Task List

• Generate caption with CNN Encoder & LSTM Decoder
• Create Web App for Inference
• Use visual Attention

Technology Stack:

PyTorch

Acknowledgement:

• Rupesh Somana For the Web Interface

Web Interface

Contributing

Code Style

Code is formatted using the black style. Merges to master are prevented if code does not conform.

To apply black to your code, run black from the root Prefect directory:

black .

Formatting can be easy to forget when developing, so you may choose to install a pre-push hook for black, as follows:

pre-commit install --hook-type pre-push

Once installed, you won't be allowed to git push without passing black.

In addition, a number of extensions are available for popular editors that will automatically apply black to your code.