AIbase
EN
Home
CLIP ViT B 32 CommonPool.M.clip S128m B4k
C

CLIP ViT B 32 CommonPool.M.clip S128m B4k

Developed by laion
Zero-shot image classification model based on CLIP architecture, supporting general pooling functionality
Image-to-TextOpen Source License:MIT#Zero-shot image classification#Multimodal contrastive learning#Large-scale pretraining
Downloads 164
Release Time : 4/26/2023

Model Overview

This model is a vision-language model based on the CLIP architecture, capable of performing zero-shot image classification tasks. It combines a Vision Transformer (ViT-B-32) and a text encoder, trained on a large number of image-text pairs through contrastive learning.

Model Features

Zero-shot learning capability
Performs image classification tasks without task-specific fine-tuning
General pooling functionality
Supports multiple pooling strategies to enhance model adaptability across different tasks
Vision-language alignment
Aligns visual and textual representations into the same space through contrastive learning

Model Capabilities

Zero-shot image classification
Image-text matching
Cross-modal retrieval

Use Cases

Content moderation
Automatic content filtering
Automatically identifies inappropriate content based on text descriptions
E-commerce
Product image classification
Automatically classifies product images based on descriptions
Media analysis
Image captioning
Generates descriptive labels for images
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
EnglishįŽ€äŊ“中文įšéĢ”中文ãĢãģんご
Š 2025AIbase