Search Results for author: Juan C. SanMiguel

Found 7 papers, 2 papers with code

Soft labelling for semantic segmentation: Bringing coherence to label down-sampling

1 code implementation27 Feb 2023 Roberto Alcover-Couso, Marcos Escudero-Vinolo, Juan C. SanMiguel, Jose M. Martinez

To that aim, we propose a novel framework for label down-sampling via soft-labeling that better conserves label information after down-sampling.

Data Augmentation Semantic Segmentation

Detection-aware multi-object tracking evaluation

no code implementations16 Dec 2022 Juan C. SanMiguel, Jorge Muñoz, Fabio Poiesi

How would you fairly evaluate two multi-object tracking algorithms (i. e. trackers), each one employing a different object detector?

Multi-Object Tracking Object

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

Graph Neural Networks for Cross-Camera Data Association

2 code implementations17 Jan 2022 Elena Luna, Juan C. SanMiguel, José M. Martínez, Pablo Carballeira

To avoid the usage of fixed distances, we leverage the connectivity of Graph Neural Networks, previously unused in this scope, using a Message Passing Network to jointly learn features and similarity.

3D Pose Estimation Graph Matching +1

Online Clustering-based Multi-Camera Vehicle Tracking in Scenarios with overlapping FOVs

no code implementations8 Feb 2021 Elena Luna, Juan C. SanMiguel, Jose M. Martínez, Marcos Escudero-Viñolo

Multi-Target Multi-Camera (MTMC) vehicle tracking is an essential task of visual traffic monitoring, one of the main research fields of Intelligent Transportation Systems.

Clustering Online Clustering

On guiding video object segmentation

no code implementations25 Apr 2019 Diego Ortego, Kevin McGuinness, Juan C. SanMiguel, Eric Arazo, José M. Martínez, Noel E. O'Connor

This guiding process relies on foreground masks from independent algorithms (i. e. state-of-the-art algorithms) to implement an attention mechanism that incorporates the spatial location of foreground and background to compute their separated representations.

Foreground Segmentation Object +5

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