Murtaza, Shakeeb, Belharbi, Soufiane, Guichemerre, Alexis, Pedersoli, Marco et Granger, Eric.
2027.
« TeD-Loc: Text distillation for weakly supervised object localization ».
Pattern Recognition, vol. 183, nº Part B.
Prévisualisation |
PDF
Pedersoli-M-2026-34685.pdf - Version publiée Licence d'utilisation : Creative Commons CC BY-NC-ND. Télécharger (4MB) | Prévisualisation |
Résumé
Weakly supervised object localization (WSOL) models can predict both the object class and the spatial regions corresponding to the object, without requiring explicit bounding-box annotations. Given their reliance on classification objectives, traditional WSOL methods, like class activation mapping, tend to focus on the most discriminative object regions, often missing the full spatial extent. Although vision–language models like CLIP encode rich semantic priors, their global text and class-token embeddings are not explicitly aligned with local patch embeddings, limiting patch-level localization. Recent methods such as GenPrompt address this limitation, but at the cost of increased complexity, as they rely on conditional denoising and elaborate prompt-learning strategies. In this paper, we propose Text Distillation for Localization (TeD-Loc), which distills knowledge from CLIP text embeddings to patch embeddings through contrastive alignment, thereby enabling patch-level foreground/background localization. A localization-guided classification module is also introduced, which uses localization scores to aggregate foreground patch embeddings for joint classification and localization within a single model. In addition, a QR-based orthogonalization of class text embeddings is applied before distillation to improve discrimination for semantically similar classes. Extensive experiments1 show that TeD-Loc improves Top-1Loc by ∼ 5% on CUB and ILSVRC, and PxAP by ∼ 31% on histopathology benchmarks, while achieving more efficient inference than GenPrompt.
| Type de document: | Article publié dans une revue, révisé par les pairs |
|---|---|
| Chercheur(-euse): | Chercheur(-euse) Pedersoli, Marco Granger, Éric |
| Affiliation: | Génie des systèmes, Génie des systèmes |
| Date de dépôt: | 28 oct. 2026 20:01 |
| Dernière modification: | 30 oct. 2026 19:47 |
| URI: | https://espace2.etsmtl.ca/id/eprint/34685 |
Actions (Authentification requise)
 |
Dernière vérification avant le dépôt |