An Adapter-free Fine-tuning Approach for Tuning 3D Foundation Models
Sneha Paul, Zachary Patterson, Nizar Bouguila · 2026
Read An Adapter-free Fine-tuning Approach for Tuning 3D Foundation Models by Sneha Paul, Zachary Patterson, Nizar Bouguila (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-23730.txt.
Description
Paul, Patterson, and Bouguila tune 3D foundation models without bolting on adapters. Point cloud giants adapt poorly on scarce data: full tuning overfits and drifts. Adapter tricks add parts that slow each run. Their momentum-consistency route tunes part of the encoder while a momentum rule guards the base skills. Two variants add unlabeled data help or prune layers for speed. Tests on shape tasks beat prior methods, with few-shot gains over three points.
Catalog record
| CREATOR | Sneha Paul, Zachary Patterson, Nizar Bouguila |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33595 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.23730