← Back to all news
arXiv Computer Vision By Inseok Kong, Geunyoung Jung, Jiyoung Jung

SATOR: Backpropagation-Free 3D Test-Time Adaptation via Sample-Adaptive Token Routing

Read the original on arXiv Computer Vision →

SATOR is a source‑free, backpropagation‑free framework for test‑time adaptation of 3D point cloud models. It eliminates the need for predefined purge‑size candidates by computing token abnormality scores from attention and geometric graphs, then using an Otsu‑based threshold to partition tokens into retained and pruned sets. A keep‑ratio‑calibrated routing step selects the pruned prediction only when its entropy, adjusted by the keep ratio, is lower than that of the original prediction, achieving state‑of‑the‑art accuracy on three 3D corruption benchmarks.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Computer Vision.

Related stories

Hugging Face Trending Papers
Jul 26

PathSelect: Sequential Token Selection for Whole Slide Pathology

Gigapixel Whole-Slide Images (WSIs) present a fundamental computational bottleneck for vision-language models (VLMs) due to extreme sequence lengths. Existing approaches predominantly rely on spatial sampling or training-free pruning, which risk diluting weak but informative signals, leading to the loss of critical diagnostic evidence due to the spatially diffuse nature of pathological cues.

llmsefficiencymultimodal
More like this →
arXiv Computer Vision
Sep 17

Position Anchor Tuning: Towards Efficient Adaptation of Pre-Trained Point Cloud Transformers

Position Anchor Tuning (PAT) is a parameter‑efficient fine‑tuning method for pre‑trained point cloud transformers that improves inference efficiency. PAT reduces the computational cost of multi‑head attention and feed‑forward networks by using token aggregation‑expansion pairs: a token aggregation module (TAM) selects representative tokens based on 3D position anchors, and a token expansion module (TEM) propagates the learned representations back to the full token set. Combined with base‑sharing low‑rank adaptation (BSLoRA) for the TAMs, PAT achieves performance comparable to state‑of‑the‑art methods while requiring fewer trainable parameters and lower computational overhead.

By Zheng Liu, Xin Gao, Jinchao Zhu, Gao Huang
llmsfine-tuningbenchmarks
More like this →