Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2511.20721 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915892116324352 |
|---|---|
| author | Letellier, Guillaume Srivastava, Siddharth Jurie, Frédéric Sharma, Gaurav |
| author_facet | Letellier, Guillaume Srivastava, Siddharth Jurie, Frédéric Sharma, Gaurav |
| contents | Foundation models pre-trained with self-supervised learning (SSL) on large-scale datasets have become powerful general-purpose feature extractors. However, their immense size and computational cost make them prohibitive for deployment on edge devices such as robots and AR/VR headsets. Existing compression techniques like standard knowledge distillation create efficient 'specialist' models but sacrifice the crucial, downstream-agnostic generality that makes foundation models so valuable. In this paper, we introduce Foundation Model Distillation (FMD), a new paradigm for compressing large SSL models into compact, efficient, and faithful proxies that retain their general-purpose representational power. We present Foundry, the first implementation of FMD for 3D point clouds. Our approach, Foundry, trains a student to learn a compressed set of SuperTokens that reconstruct the teacher's token-level representations, capturing a compact basis of its latent space. A single distilled model maintains strong transferability across diverse downstream tasks-classification, part segmentation, and few-shot scenarios-approaching full foundation-model performance while using significantly fewer tokens and FLOPs, making such models more practical for deployment on resourceconstrained hardware. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_20721 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Foundry: Distilling 3D Foundation Models for the Edge Letellier, Guillaume Srivastava, Siddharth Jurie, Frédéric Sharma, Gaurav Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Neural and Evolutionary Computing Foundation models pre-trained with self-supervised learning (SSL) on large-scale datasets have become powerful general-purpose feature extractors. However, their immense size and computational cost make them prohibitive for deployment on edge devices such as robots and AR/VR headsets. Existing compression techniques like standard knowledge distillation create efficient 'specialist' models but sacrifice the crucial, downstream-agnostic generality that makes foundation models so valuable. In this paper, we introduce Foundation Model Distillation (FMD), a new paradigm for compressing large SSL models into compact, efficient, and faithful proxies that retain their general-purpose representational power. We present Foundry, the first implementation of FMD for 3D point clouds. Our approach, Foundry, trains a student to learn a compressed set of SuperTokens that reconstruct the teacher's token-level representations, capturing a compact basis of its latent space. A single distilled model maintains strong transferability across diverse downstream tasks-classification, part segmentation, and few-shot scenarios-approaching full foundation-model performance while using significantly fewer tokens and FLOPs, making such models more practical for deployment on resourceconstrained hardware. |
| title | Foundry: Distilling 3D Foundation Models for the Edge |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning Neural and Evolutionary Computing |
| url | https://arxiv.org/abs/2511.20721 |