Nov 13 – 14, 2024
Europe/Berlin timezone

Can Biases in ImageNet Models Explain Generalization?

Nov 14, 2024, 2:00 PM
45m
Nectar Poster NECTAR Main Conference

Description

The robust generalization of models to rare, in-distribution (ID) samples drawn from the long tail of the training distribution and to out-of-training-distribution (OOD) samples is one of the major challenges of current deep learning methods. For image classification, this manifests in the existence of adversarial attacks, the performance drops on distorted images, and a lack of generalization to concepts such as sketches. The current understanding of generalization in neural networks is very limited, but some biases that differentiate models from human vision have been identified and might be causing these limitations. Consequently, several attempts with varying success have been made to reduce these biases during training to improve generalization. We take a step back and sanity-check these attempts. Fixing the architecture to the well-established ResNet-50, we perform a large-scale study on 48 ImageNet models obtained via different training methods to understand how and if these biases - including shape bias, spectral biases, and critical bands - interact with generalization. Our extensive study results reveal that contrary to previous findings, these biases are insufficient to accurately predict the generalization of a model holistically. We provide access to all checkpoints and evaluation code at https://github.com/paulgavrikov/biases_vs_generalization

URL of original paper (Nectar only) https://arxiv.org/abs/2404.01509
Original Venue (Nectar only) CVPR 2024

Primary authors

Paul Gavrikov (IMLA) Janis Keuper (Fakultät EMI)

Presentation materials

There are no materials yet.