Bongard-HOI

Dataset Information
Modalities
Images, Texts
Languages
English
Introduced
2022
License
CC-BY-4.0
Homepage
Official Website
Contents

Overview

Bongard-HOI testifies to which extent your few-shot visual learner can quickly induce the true HOI concept from a handful of images and perform reasoning with it. Further, the learner is also expected to transfer the learned few-shot skills to novel HOI concepts compositionally.

Variants: Bongard-HOI

Associated Benchmarks

This dataset is used in 1 benchmark:

Recent Benchmark Submissions

Task Model Paper Date
Few-Shot Image Classification GPT-4o + CA A Cognitive Paradigm Approach to … 2025-01-23
Few-Shot Image Classification Gemini 2.0 + CA A Cognitive Paradigm Approach to … 2025-01-23
Few-Shot Image Classification SVM-Mimic + PMF (fine-tuned CLIP RN-50) Support-Set Context Matters for Bongard … 2023-09-07
Few-Shot Image Classification SVM-Mimic (frozen CLIP RN-50) Support-Set Context Matters for Bongard … 2023-09-07
Few-Shot Image Classification Human (Amateur) Bongard-HOI: Benchmarking Few-Shot Visual Reasoning … 2022-05-27
Few-Shot Image Classification ANIL (ImageNet_R50) Bongard-HOI: Benchmarking Few-Shot Visual Reasoning … 2022-05-27
Few-Shot Image Classification Meta-Baseline (ImagNet_R50) Bongard-HOI: Benchmarking Few-Shot Visual Reasoning … 2022-05-27
Few-Shot Image Classification Meta-Baseline (MoCov2_R50) Bongard-HOI: Benchmarking Few-Shot Visual Reasoning … 2022-05-27
Few-Shot Image Classification Meta-Baseline (Scratch_R50) Bongard-HOI: Benchmarking Few-Shot Visual Reasoning … 2022-05-27

Research Papers

Recent papers with results on this dataset:

External Links: