Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Knobel, Lukas, Han, Tengda, Asano, Yuki M.
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2307.08727
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914733736591360
author Knobel, Lukas
Han, Tengda
Asano, Yuki M.
author_facet Knobel, Lukas
Han, Tengda
Asano, Yuki M.
contents While recent supervised methods for reference-based object counting continue to improve the performance on benchmark datasets, they have to rely on small datasets due to the cost associated with manually annotating dozens of objects in images. We propose UnCounTR, a model that can learn this task without requiring any manual annotations. To this end, we construct "Self-Collages", images with various pasted objects as training samples, that provide a rich learning signal covering arbitrary object types and counts. Our method builds on existing unsupervised representations and segmentation techniques to successfully demonstrate for the first time the ability of reference-based counting without manual supervision. Our experiments show that our method not only outperforms simple baselines and generic models such as FasterRCNN and DETR, but also matches the performance of supervised counting models in some domains.
format Preprint
id arxiv_https___arxiv_org_abs_2307_08727
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Learning to Count without Annotations
Knobel, Lukas
Han, Tengda
Asano, Yuki M.
Computer Vision and Pattern Recognition
While recent supervised methods for reference-based object counting continue to improve the performance on benchmark datasets, they have to rely on small datasets due to the cost associated with manually annotating dozens of objects in images. We propose UnCounTR, a model that can learn this task without requiring any manual annotations. To this end, we construct "Self-Collages", images with various pasted objects as training samples, that provide a rich learning signal covering arbitrary object types and counts. Our method builds on existing unsupervised representations and segmentation techniques to successfully demonstrate for the first time the ability of reference-based counting without manual supervision. Our experiments show that our method not only outperforms simple baselines and generic models such as FasterRCNN and DETR, but also matches the performance of supervised counting models in some domains.
title Learning to Count without Annotations
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2307.08727