CeRVIM Seminar: Akshaya Athwale, February 16, 2024

CeRVIM Seminar: DarSwin: Distortion-Aware Radial Swin Transformers for Wide Angle Image Recognition

Akshaya Athwale
Laboratoire de Vision et Systèmes Numériques, LVSN
Dép. de génie électrique et de génie informatique, U. Laval

Friday, February 16, 2024, 11:00 a.m., PLT-3370

Wide-angle lenses are commonly used in perception tasks requiring a large field of view. Unfortunately, these lenses produce significant distortions, making conventional models that ignore the distortion effects unable to adapt to wide-angle images. In this research, we present a novel transformer-based model that automatically adapts to the distortion produced by wide-angle lenses. Our proposed image encoder architecture, dubbed DarSwin, leverages the physical characteristics of such lenses analytically defined by the radial distortion profile. In contrast to conventional transformer-based architectures, DarSwin comprises a radial patch partitioning, a distortion-based sampling technique for creating token embeddings, and an angular position encoding for radial patch merging. Compared to other baselines, DarSwin achieves the best results on different datasets with significant gains when trained on bounded levels of distortions (very low, low, medium, and high) and tested on all, including out-of-distribution distortions. While the base DarSwin architecture requires knowledge of the radial distortion profile, we show it can be combined with a self-calibration network that estimates such a profile from the input image itself, resulting in a completely uncalibrated pipeline. Finally, we also present DarSwin-Unet, which extends DarSwin to an encoder-decoder architecture suitable for pixel-level tasks. We demonstrate its performance on depth estimation and show through extensive experiments that DarSwin-Unet can perform zero-shot adaptation to unseen distortions of different wide-angle lenses.

The presentation will be given in English and the slides will be in English.