Category Ranking
98%
Total Visits
921
Avg Visit Duration
2 minutes
Citations
20
Article Abstract
Advances in computer vision and increasingly widespread video-based behavioral monitoring are currently transforming how we study animal behavior. However, there is still a gap between the prospects and practical application, especially in videos from the wild. In this Perspective, we aim to present the capabilities of current methods for behavioral analysis, while at the same time highlighting unsolved computer vision problems that are relevant to the study of animal behavior. We survey state-of-the-art methods for computer vision problems relevant to the video-based study of individualized animal behavior, including object detection, multi-animal tracking, individual identification and (inter)action understanding. We then review methods for effort-efficient learning, one of the challenges from a practical perspective. In our outlook on the emerging field of computer vision for animal behavior, we argue that the field should develop approaches to unify detection, tracking, identification and (inter)action understanding in a single, video-based framework.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1038/s41592-025-02653-y | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Deep feature engineering for accurate sperm morphology classification using CBAM-enhanced ResNet50.
PLoS One
September 2025
School of Computer Science, CHART Laboratory, University of Nottingham, Nottingham, United Kingdom.
Background And Objective: Male fertility assessment through sperm morphology analysis remains a critical component of reproductive health evaluation, as abnormal sperm morphology is strongly correlated with reduced fertility rates and poor assisted reproductive technology outcomes. Traditional manual analysis performed by embryologists is time-intensive, subjective, and prone to significant inter-observer variability, with studies reporting up to 40% disagreement between expert evaluators. This research presents a novel deep learning framework combining Convolutional Block Attention Module (CBAM) with ResNet50 architecture and advanced deep feature engineering (DFE) techniques for automated, objective sperm morphology classification.
View Article and Find Full Text PDFMeasuring Street Built Environments for Children's Use: A Systematic Review of Measurement Tools.
J Urban Health
September 2025
School of Architecture and Design, Harbin Institute of Technology, Harbin, 150001, China.
Street-level environments play a vital role in children's development by promoting their physical activity, cognitive growth, and overall development. This study systematically reviews the measurement tools available to assess street environments according to children's needs. This systematic review was conducted according to the PRISMA-COSMIN guidelines.
View Article and Find Full Text PDFDecoding binocular color differences via EEG signals: linking ERP dynamics to chromatic disparity in CIELAB space.
Exp Brain Res
September 2025
School of Information Science and Technology, Yunnan Normal University, Kunming, 650500, China.
This study explores how differences in colors presented separately to each eye (binocular color differences) can be identified through EEG signals, a method of recording electrical activity from the brain. Four distinct levels of green-red color differences, defined in the CIELAB color space with constant luminance and chroma, are investigated in this study. Analysis of Event-Related Potentials (ERPs) revealed a significant decrease in the amplitude of the P300 component as binocular color differences increased, suggesting a measurable brain response to these differences.
View Article and Find Full Text PDFMultimodal machine learning for staging laparoscopy: a combined image analysis and morphologic tool for the discrimination of peritoneal metastasis.
Int J Surg
September 2025
Department of Human Structure and Repair, Ghent University Faculty of Medicine, Belgium.
Background: Staging laparoscopy (SL) is an essential procedure for peritoneal metastasis (PM) detection. Although surgeons are expected to differentiate between benign and malignant lesions intraoperatively, this task remains difficult and error-prone. The aim of this study was to develop a novel multimodal machine learning (MML) model to differentiate PM from benign lesions by integrating morphologic characteristics with intraoperative SL images.
View Article and Find Full Text PDFReducing motion artifacts in craniocervical background subtraction angiography with deformable registration and unsupervised deep learning.
Radiol Adv
September 2024
Department of Radiology, Northwestern University and Northwestern Medicine, Chicago, IL, 60611, United States.
Background: In clinical practice, digital subtraction angiography (DSA) often suffers from misregistration artifact resulting from voluntary, respiratory, and cardiac motion during acquisition. Most prior efforts to register the background DSA mask to subsequent postcontrast images rely on key point registration using iterative optimization, which has limited real-time application.
Purpose: Leveraging state-of-the-art, unsupervised deep learning, we aim to develop a fast, deformable registration model to substantially reduce DSA misregistration in craniocervical angiography without compromising spatial resolution or introducing new artifacts.