ExploreCfg
by yongliang-wu
[NeurIPS2023] Exploring Diverse In-Context Configurations for Image Captioning
AI summary
Config strategy development
This project develops strategies to optimize in-context sequence configurations for Vision-Language few-shot learning, with a focus on exploring the effects of varying configurations on image-text pairs.
- stars
- 33
- forks
- 0
- watching
- 0
Similar projects
Found by comparing what the projects do, not just their names.
Image Captioner
A PyTorch implementation of image captioning models via scene graph decomposition.
Image captioning model
An image caption generation model that uses abstract scene graphs to fine-grained control and generate captions
Image captioner
An unsupervised image captioning framework that allows generating captions from images without paired data.
configurator
Provides configuration support for a dependency injection container
Layout generator
This project generates visual layouts and scenes using large language models.
Image captioner
Enhances language models to generate text based on visual descriptions of images
go-gcfg/gcfg168
Config parser
A Go library that reads INI-style configuration files into data structures
Scene graph generator
Automatically generates scene graphs from images to aid in image captioning tasks
Video captioner
PyTorch implementation of video captioning, combining deep learning and computer vision techniques.
jkcfg/jk406
Configurer
A tool for writing structured configuration files using a general-purpose language
Image generator
An image caption generation system utilizing machine learning models and deep neural networks.
Image captioning model
A deep learning model for generating image captions with semantic attention
Video captioning model
An implementation of a dense video captioning model with attention-based fusion and context gating
Caption generator
Trains image paragraph captioning models to generate diverse and accurate captions
Caption generator
A PyTorch implementation of a framework for generating captions with styles for images and videos.
