Test-time Adaptation for Point Cloud Denoising under Unseen Noise

Key points

• Although deep learning-based point cloud denoising methods have demonstrated impressive results in recent years, their effectiveness usually degrades substantially in real-world scenarios where the noise distribution of test point clouds deviates from that of the training set.1
• Taking a pre-trained denoising model and a test noisy point cloud as input, TTAD first employs a residual-driven noise modeling (RDNM) module to estimate the sample-specific noise distribution through analyzing the deviation between the noisy input and its pseudo-clean version denoised by the initial model.1
• Then, a noise-aware contrastive adaptation (NACA) module is designed to build sample-specific noisy-noisier test-time training pair from the denoised point cloud, and update the model parameters over a few adaptation steps per sample at test time through contrastive learning.1
• Extensive experiments on both synthetic and real-world datasets demonstrate that our approach significantly improves the generalization ability of existing denoising models under unseen noise, achieving superior performance across multiple evaluation metrics.1

Sources (1)

• [1]Test-time Adaptation for Point Cloud Denoising under Unseen Noise
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 20, 12:04 AM

  To address this challenge, we propose a test-time adaptive denoising (TTAD) framework that enables existing denoising approaches to adapt to unknown noise patterns at test time without relying on additional supervision or retraining from source data.

  Although deep learning-based point cloud denoising methods have demonstrated impressive results in recent years, their effectiveness usually degrades substantially in real-world scenarios where the noise distribution of test point clouds deviates from that of the training set.

Extractive summary: sentences quoted from the sources.