Salvato in:
Dettagli Bibliografici
Autori principali: Tourani, Siddharth, Alwheibi, Ahmed, Mahmood, Arif, Khan, Muhammad Haris
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:https://arxiv.org/abs/2403.16194
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911811814555648
author Tourani, Siddharth
Alwheibi, Ahmed
Mahmood, Arif
Khan, Muhammad Haris
author_facet Tourani, Siddharth
Alwheibi, Ahmed
Mahmood, Arif
Khan, Muhammad Haris
contents Unsupervised landmarks discovery (ULD) for an object category is a challenging computer vision problem. In pursuit of developing a robust ULD framework, we explore the potential of a recent paradigm of self-supervised learning algorithms, known as diffusion models. Some recent works have shown that these models implicitly contain important correspondence cues. Towards harnessing the potential of diffusion models for the ULD task, we make the following core contributions. First, we propose a ZeroShot ULD baseline based on simple clustering of random pixel locations with nearest neighbour matching. It delivers better results than existing ULD methods. Second, motivated by the ZeroShot performance, we develop a ULD algorithm based on diffusion features using self-training and clustering which also outperforms prior methods by notable margins. Third, we introduce a new proxy task based on generating latent pose codes and also propose a two-stage clustering mechanism to facilitate effective pseudo-labeling, resulting in a significant performance improvement. Overall, our approach consistently outperforms state-of-the-art methods on four challenging benchmarks AFLW, MAFL, CatHeads and LS3D by significant margins.
format Preprint
id arxiv_https___arxiv_org_abs_2403_16194
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Pose-Guided Self-Training with Two-Stage Clustering for Unsupervised Landmark Discovery
Tourani, Siddharth
Alwheibi, Ahmed
Mahmood, Arif
Khan, Muhammad Haris
Computer Vision and Pattern Recognition
Unsupervised landmarks discovery (ULD) for an object category is a challenging computer vision problem. In pursuit of developing a robust ULD framework, we explore the potential of a recent paradigm of self-supervised learning algorithms, known as diffusion models. Some recent works have shown that these models implicitly contain important correspondence cues. Towards harnessing the potential of diffusion models for the ULD task, we make the following core contributions. First, we propose a ZeroShot ULD baseline based on simple clustering of random pixel locations with nearest neighbour matching. It delivers better results than existing ULD methods. Second, motivated by the ZeroShot performance, we develop a ULD algorithm based on diffusion features using self-training and clustering which also outperforms prior methods by notable margins. Third, we introduce a new proxy task based on generating latent pose codes and also propose a two-stage clustering mechanism to facilitate effective pseudo-labeling, resulting in a significant performance improvement. Overall, our approach consistently outperforms state-of-the-art methods on four challenging benchmarks AFLW, MAFL, CatHeads and LS3D by significant margins.
title Pose-Guided Self-Training with Two-Stage Clustering for Unsupervised Landmark Discovery
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.16194