facebookresearch/ImageBind

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

9064 stars · 843 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

A PyTorch library from Meta AI implementing ImageBind, a model that learns a joint embedding space across six modalities: images, text, audio, depth, thermal, and IMU data. It ships pretrained checkpoints and enables cross-modal retrieval, zero-shot classification, and modality arithmetic out of the box.

Use cases

• embed images and text into a shared vector space
• cross-modal retrieval like finding images from audio or text
• zero-shot classification across modalities
• compose modalities with embedding arithmetic
• generate embeddings for depth, thermal, or IMU sensor data
• build multimodal search or RAG pipelines with unified embeddings

When to avoid

• you only need standard image-text embeddings like CLIP with lighter dependencies
• you need production-grade inference serving rather than a research codebase
• you need modalities outside the six supported ones
• you require a permissive license - the license is custom, not standard open source

Facets

library · maturity maintenance

machine-learning deep-learning image-processing audio-processing nlp deep-learning machine-learning artificial-intelligence computer-vision python windows multimodal-embeddings joint-embedding-space cross-modal-retrieval pytorch pretrained-models zero-shot-classification research-model audio gpu linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="facebookresearch/ImageBind")