TeD-Loc: Text Distillation for Weakly Supervised Object Localization

arXiv:2501.12632v2 Announce Type: replace-cross Abstract: Weakly supervised object localization (WSOL) models are trained using only image-level class labels. They can predict both the object class and 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 such as CLIP encode rich semantic priors, they are not directly suited for WSOL because global text and class-token embeddings are not explicitly aligned with local patch embeddings, making patch-level localization difficult without additional mechanisms. 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. We propose Text Distillation for Localization (TeD-Loc), which transfers 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 that uses localization scores to aggregate foreground patch embeddings for joint classification and localization in 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 experiments show that TeD-Loc improves Top-1 Loc by ~5% on CUB and ILSVRC, and PxAP by ~31% on histopathology benchmarks, while achieving more efficient inference than GenPrompt.