Home
Information

AI Dataset Collection

Large-scale datasets and benchmarks for training, evaluating, and testing models to measure

Tools

Intelligent Document Recognition

Comprehensive Text Extraction and Document Processing Solutions for Users

AI Tutorial

blip-vision-language

Public

BLIP is a novel Vision-Language Pre-training (VLP) framework designed to excel in both understanding and generation tasks by salesforce. Unlike existing models, BLIP leverages noisy web data through bootstrapping, generating synthetic captions and effectively filtering out noise for enhanced performance.

Creat2023-08-22T23:25:26
Update2024-05-24T02:05:04
3
Stars
0
Stars Increase

Related projects