Search Results for author: Jesús Bescós

Found 5 papers, 2 papers with code

Attention-based Knowledge Distillation in Multi-attention Tasks: The Impact of a DCT-driven Loss

no code implementations4 May 2022 Alejandro López-Cifuentes, Marcos Escudero-Viñolo, Jesús Bescós, Juan C. SanMiguel

Feature-based Knowledge Distillation is a subfield of KD that relies on intermediate network representations, either unaltered or depth-reduced via maximum activation maps, as the source knowledge.

Descriptive Knowledge Distillation +1

Semantic Driven Multi-Camera Pedestrian Detection

no code implementations27 Dec 2018 Alejandro López-Cifuentes, Marcos Escudero-Viñolo, Jesús Bescós, Pablo Carballeira

Contrarily to the majority of the methods of the state-of-the-art, the proposed approach is scene-agnostic, not requiring a tailored adaptation to the target scenario\textemdash e. g., via fine-tunning.

Attribute Pedestrian Detection +1

Cannot find the paper you are looking for? You can Submit a new open access paper.