Localize to Classify and Classify to Localize: Mutual Guidance in Object Detection - Inria - Institut national de recherche en sciences et technologies du numérique
Conference Papers Year : 2021

Localize to Classify and Classify to Localize: Mutual Guidance in Object Detection

Abstract

Most deep learning object detectors are based on the anchor mechanism and resort to the Intersection over Union (IoU) between predefined anchor boxes and ground truth boxes to evaluate the matching quality between anchors and objects. In this paper, we question this use of IoU and propose a new anchor matching criterion guided, during the training phase, by the optimization of both the localization and the classification tasks: the predictions related to one task are used to dynamically assign sample anchors and improve the model on the other task, and vice versa. Despite the simplicity of the proposed method, our experiments with different state-of-the-art deep learning architectures on PASCAL VOC and MS COCO datasets demonstrate the effectiveness and generality of our Mutual Guidance strategy.
Fichier principal
Vignette du fichier
accv2020.pdf (9.96 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03025756 , version 1 (26-11-2020)

Identifiers

Cite

Heng Zhang, Elisa Fromont, Sébastien Lefèvre, Bruno Avignon. Localize to Classify and Classify to Localize: Mutual Guidance in Object Detection. ACCV 2020 - The 15th Asian Conference on Computer Vision, Nov 2020, Kyoto / Virtual, Japan. pp.1-15, ⟨10.1007/978-3-030-69538-5_7⟩. ⟨hal-03025756⟩
92 View
56 Download

Altmetric

Share

More