openvision-vit-base-patch16-160 Open-source Visual Encoder - Supports Multimodal Learning and Cost-effective

Openvision Vit Base Patch16 160

Developed by UCSC-VLAA

OpenVision is a fully open-source, cost-effective family of advanced vision encoders for multimodal learning.

Downloads 15

Release Time : 5/6/2025

Model Overview

OpenVision aims to provide an open-source vision encoder solution supporting multimodal learning tasks with high efficiency and cost-effectiveness.

Fully Open-source

Model code and weights are fully open-source, facilitating research and commercial use.

Cost-effective

Designed with computational efficiency and cost-effectiveness in mind, suitable for resource-constrained environments.

Multimodal Learning Support

Supports joint learning of vision and language multimodal tasks.

Image feature extraction

Multimodal learning

Vision encoding

Computer Vision

Image Classification

Use extracted image features for classification tasks.

Image Retrieval

Retrieve images based on feature similarity.

Multimodal Learning

Image-Text Matching

Map images and text to the same feature space for matching.

Property	Details
Pipeline Tag	Image Feature Extraction
Library Name	openclip

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base