Nuisance-Label Supervision: Robustness Improvement by Free Labels

10/14/2021
by   Xinyue Wei, et al.
0

In this paper, we present a Nuisance-label Supervision (NLS) module, which can make models more robust to nuisance factor variations. Nuisance factors are those irrelevant to a task, and an ideal model should be invariant to them. For example, an activity recognition model should perform consistently regardless of the change of clothes and background. But our experiments show existing models are far from this capability. So we explicitly supervise a model with nuisance labels to make extracted features less dependent on nuisance factors. Although the values of nuisance factors are rarely annotated, we demonstrate that besides existing annotations, nuisance labels can be acquired freely from data augmentation and synthetic data. Experiments show consistent improvement in robustness towards image corruption and appearance change in action recognition.

READ FULL TEXT

page 4

page 5

research
12/03/2019

RSA: Randomized Simulation as Augmentation for Robust Human Action Recognition

Despite the rapid growth in datasets for video activity, stable robust a...
research
11/03/2022

ImageNet-X: Understanding Model Mistakes with Factor of Variation Annotations

Deep learning vision systems are widely deployed across applications whe...
research
03/30/2021

Learning Representational Invariances for Data-Efficient Action Recognition

Data augmentation is a ubiquitous technique for improving image classifi...
research
09/22/2020

On Data Augmentation for Extreme Multi-label Classification

In this paper, we focus on data augmentation for the extreme multi-label...
research
12/07/2020

VideoMix: Rethinking Data Augmentation for Video Classification

State-of-the-art video action classifiers often suffer from overfitting....
research
10/04/2022

Nuisances via Negativa: Adjusting for Spurious Correlations via Data Augmentation

There exist features that are related to the label in the same way acros...
research
05/24/2019

Implicit Label Augmentation on Partially Annotated Clips via Temporally-Adaptive Features Learning

Partially annotated clips contain rich temporal contexts that can comple...

Please sign up or login with your details

Forgot password? Click here to reset