NLP course - Image Caption Generator

1. General Images

Available Datasets

MS COCO

Flickr8K

Flickr30K

Karpathy splits for Image Captioning

Existing Papers and Repos

Show, Attend and Tell: Neural Image Caption Generation with Visual Attention (code), (paper), (blog)

Show and Tell: A Neural Image Caption Generator (code), (paper)

Deep Visual-Semantic Alignments for Generating Image Descriptions (code), (paper)

Related Blogs

CNN-LSTM Architecture

图像字幕生成概述

Model

Encoder: Resnet50 (Contrastive learning)

Decoder: LSTM or Transformer

2. Chemical Images

Available Datasets

Synthetic data

Existing Repos

DECIMER-Image-to-SMILES