Make me an Expert: Distilling from Generalist Black-Box Models into Specialized Models for Semantic Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Benigmim, Yasser, Roy, Subhankar, Oublal, Khalid, Marouf, Imad Eddine, Essid, Slim, Kalogeiton, Vicky, Lathuilière, Stéphane
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912562121015296
author Benigmim, Yasser
Roy, Subhankar
Oublal, Khalid
Marouf, Imad Eddine
Essid, Slim
Kalogeiton, Vicky
Lathuilière, Stéphane
author_facet Benigmim, Yasser
Roy, Subhankar
Oublal, Khalid
Marouf, Imad Eddine
Essid, Slim
Kalogeiton, Vicky
Lathuilière, Stéphane
contents The rise of Artificial Intelligence as a Service (AIaaS) democratizes access to pre-trained models via Application Programming Interfaces (APIs), but also raises a fundamental question: how can local models be effectively trained using black-box models that do not expose their weights, training data, or logits, a constraint in which current domain adaptation paradigms are impractical ? To address this challenge, we introduce the Black-Box Distillation (B2D) setting, which enables local model adaptation under realistic constraints: (1) the API model is open-vocabulary and trained on large-scale general-purpose data, and (2) access is limited to one-hot predictions only. We identify that open-vocabulary models exhibit significant sensitivity to input resolution, with different object classes being segmented optimally at different scales, a limitation termed the "curse of resolution". Our method, ATtention-Guided sCaler (ATGC), addresses this challenge by leveraging DINOv2 attention maps to dynamically select optimal scales for black-box model inference. ATGC scores the attention maps with entropy to identify informative scales for pseudo-labelling, enabling effective distillation. Experiments demonstrate substantial improvements under black-box supervision across multiple datasets while requiring only one-hot API predictions. Our code is available at https://github.com/yasserben/ATGC.
format Preprint
id arxiv_https___arxiv_org_abs_2509_00509
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Make me an Expert: Distilling from Generalist Black-Box Models into Specialized Models for Semantic Segmentation
Benigmim, Yasser
Roy, Subhankar
Oublal, Khalid
Marouf, Imad Eddine
Essid, Slim
Kalogeiton, Vicky
Lathuilière, Stéphane
Computer Vision and Pattern Recognition
The rise of Artificial Intelligence as a Service (AIaaS) democratizes access to pre-trained models via Application Programming Interfaces (APIs), but also raises a fundamental question: how can local models be effectively trained using black-box models that do not expose their weights, training data, or logits, a constraint in which current domain adaptation paradigms are impractical ? To address this challenge, we introduce the Black-Box Distillation (B2D) setting, which enables local model adaptation under realistic constraints: (1) the API model is open-vocabulary and trained on large-scale general-purpose data, and (2) access is limited to one-hot predictions only. We identify that open-vocabulary models exhibit significant sensitivity to input resolution, with different object classes being segmented optimally at different scales, a limitation termed the "curse of resolution". Our method, ATtention-Guided sCaler (ATGC), addresses this challenge by leveraging DINOv2 attention maps to dynamically select optimal scales for black-box model inference. ATGC scores the attention maps with entropy to identify informative scales for pseudo-labelling, enabling effective distillation. Experiments demonstrate substantial improvements under black-box supervision across multiple datasets while requiring only one-hot API predictions. Our code is available at https://github.com/yasserben/ATGC.
title Make me an Expert: Distilling from Generalist Black-Box Models into Specialized Models for Semantic Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.00509