Webly Supervised Concept Expansion for General Purpose Vision Models

02/04/2022
by   Amita Kamath, et al.
0

General purpose vision (GPV) systems are models that are designed to solve a wide array of visual tasks without requiring architectural changes. Today, GPVs primarily learn both skills and concepts from large fully supervised datasets. Scaling GPVs to tens of thousands of concepts by acquiring data to learn each concept for every skill quickly becomes prohibitive. This work presents an effective and inexpensive alternative: learn skills from fully supervised datasets, learn concepts from web image search results, and leverage a key characteristic of GPVs – the ability to transfer visual knowledge across skills. We use a dataset of 1M+ images spanning 10k+ visual concepts to demonstrate webly-supervised concept expansion for two existing GPVs (GPV-1 and VL-T5) on 3 benchmarks - 5 COCO based datasets (80 primary concepts), a newly curated series of 5 datasets based on the OpenImages and VisualGenome repositories ( 500 concepts) and the Web-derived dataset (10k+ concepts). We also propose a new architecture, GPV-2 that supports a variety of tasks – from vision tasks like classification and localization to vision+language tasks like QA and captioning to more niche ones like human-object interaction recognition. GPV-2 benefits hugely from web data, outperforms GPV-1 and VL-T5 across these benchmarks, and does well in a 0-shot setting at action and attribute recognition.

READ FULL TEXT

page 1

page 4

page 7

page 14

page 15

page 16

page 17

research
04/01/2021

Towards General Purpose Vision Systems

A special purpose learning system assumes knowledge of admissible tasks ...
research
04/28/2022

GRIT: General Robust Image Task Benchmark

Computer vision models excel at making predictions when the test distrib...
research
05/23/2023

Can Language Models Understand Physical Concepts?

Language models (LMs) gradually become general-purpose interfaces in the...
research
11/17/2022

I Can't Believe There's No Images! Learning Visual Tasks Using only Language Data

Many high-level skills that are required for computer vision tasks, such...
research
10/12/2021

Can machines learn to see without visual databases?

This paper sustains the position that the time has come for thinking of ...
research
06/09/2017

Learning to Learn from Noisy Web Videos

Understanding the simultaneously very diverse and intricately fine-grain...
research
06/03/2022

Drawing out of Distribution with Neuro-Symbolic Generative Models

Learning general-purpose representations from perceptual inputs is a hal...

Please sign up or login with your details

Forgot password? Click here to reset