Angular Gap: Reducing the Uncertainty of Image Difficulty through Model Calibration

07/18/2022
by   Bohua Peng, et al.
0

Curriculum learning needs example difficulty to proceed from easy to hard. However, the credibility of image difficulty is rarely investigated, which can seriously affect the effectiveness of curricula. In this work, we propose Angular Gap, a measure of difficulty based on the difference in angular distance between feature embeddings and class-weight embeddings built by hyperspherical learning. To ascertain difficulty estimation, we introduce class-wise model calibration, as a post-training technique, to the learnt hyperbolic space. This bridges the gap between probabilistic model calibration and angular distance estimation of hyperspherical learning. We show the superiority of our calibrated Angular Gap over recent difficulty metrics on CIFAR10-H and ImageNetV2. We further propose Angular Gap based curriculum learning for unsupervised domain adaptation that can translate from learning easy samples to mining hard samples. We combine this curriculum with a state-of-the-art self-training method, Cycle Self Training (CST). The proposed Curricular CST learns robust representations and outperforms recent baselines on Office31 and VisDA 2017.

READ FULL TEXT

page 12

page 13

research
10/30/2020

Dynamic Data Selection for Curriculum Learning via Ability Estimation

Curriculum learning methods typically rely on heuristics to estimate the...
research
02/09/2023

Mixed-order self-paced curriculum learning for universal lesion detection

Self-paced curriculum learning (SCL) has demonstrated its great potentia...
research
07/02/2020

Curriculum Manager for Source Selection in Multi-Source Domain Adaptation

The performance of Multi-Source Unsupervised Domain Adaptation depends s...
research
07/14/2023

HuCurl: Human-induced Curriculum Discovery

We introduce the problem of curriculum discovery and describe a curricul...
research
07/03/2020

A Competence-aware Curriculum for Visual Concepts Learning via Question Answering

Humans can progressively learn visual concepts from easy to hard questio...
research
12/15/2022

Curriculum Learning Meets Weakly Supervised Modality Correlation Learning

In the field of multimodal sentiment analysis (MSA), a few studies have ...
research
08/24/2021

Density-Based Dynamic Curriculum Learning for Intent Detection

Pre-trained language models have achieved noticeable performance on the ...

Please sign up or login with your details

Forgot password? Click here to reset