dali-does/vse-probing

Code for COLING2020 paper: Probing Multimodal Embeddings for Linguistic Properties.

21
/ 100
Experimental

This tool helps researchers analyze how well multimodal AI models understand linguistic concepts in images and text. It takes existing image-caption datasets (like MSCOCO) and pretrained visual-semantic embedding models as input, then runs tests to see if the models have learned properties like object categories or semantic congruence. Researchers in AI and natural language processing can use this to evaluate and compare the linguistic capabilities of different multimodal models.

No commits in the last 6 months.

Use this if you are an AI researcher wanting to understand the linguistic knowledge captured within visual-semantic embedding models.

Not ideal if you are looking to train or deploy a new visual-semantic model, as this tool focuses on analyzing existing ones.

AI-research natural-language-processing computer-vision multimodal-AI model-analysis
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

9

Forks

Language

Python

License

Apache-2.0

Last pushed

Apr 12, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/embeddings/dali-does/vse-probing"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.