Results 81 to 90 of about 11,426 (251)

Inferring Brown Bear Hair Snare Interactions by Automatically Detecting Bipedalism on Camera Trap Images Using Pose Estimation and a Multilayer Perceptron

Remote Sensing in Ecology and Conservation, EarlyView.
This study proposes an automated method to infer brown bear hair snare interactions by detecting bipedal behavior in camera‐trap images using a pose estimation model and a multilayer perceptron (MLP). A YOLO‐based model, fine‐tuned from humans and dogs to a custom dataset, achieved high performance (≈93% keypoint precision and ≈96% classification ...
Arnau Campanera‐Moliné, Víctor Sazatornil, Lluís Brotons, Felipe Lumbreras Ruiz, Robert Benavente, Maguiña Ramilo‐Henry, Cristina Ramos, Santiago Palazón, Daniel Ponsa   +8 more
wiley   +1 more source

Boosting Binary Keypoint Descriptors [PDF]

2013 IEEE Conference on Computer Vision and Pattern Recognition, 2013
Binary key point descriptors provide an efficient alternative to their floating-point competitors as they enable faster processing while requiring less memory. In this paper, we propose a novel framework to learn an extremely compact binary descriptor we call Bin Boost that is very robust to illumination and viewpoint changes.
Tomasz Trzcinski, C. Mario Christoudias, Pascal Fua, Vincent Lepetit   +3 more
openaire   +1 more source

Open-Vocabulary Animal Keypoint Detection with Semantic-feature Matching [PDF]

Current image-based keypoint detection methods for animal (including human) bodies and faces are generally divided into full-supervised and few-shot class-agnostic approaches. The former typically relies on laborious and time-consuming manual annotations,
Luo, Ping, Shao, Wenqi, Qiao, Yu, Zhang, Kaipeng, Xu, Lumin, Zheng, Nanning, Zhang, Hao, Lai, Shenqi   +7 more
core   +1 more source

Fusion or Confusion? Assessing the Impact of Visible‐Thermal Image Fusion for Automated Wildlife Detection

Remote Sensing in Ecology and Conservation, EarlyView.
This study evaluates the performances of synchronous aerial visible (VIS) and thermal infrared (TIR) imagery for detecting great blue heron (Ardea herodias) nests and individuals using a YOLO11n model. VIS and TIR images were automatically aligned using deep learning, and both early and late fusion approaches were tested.
Camille Dionne‐Pierre, Samuel Foucher, Jérôme Théau, Jérôme Lemaître, Patrick Charbonneau, Maxime Brousseau, Mathieu Varin   +6 more
wiley   +1 more source

Fast Feature-Based Template Matching, Based on Efficient Keypoint Extraction

, 2011
In order to improve the performance of feature-based template matching techniques, several research papers have been published. Real-time applications require the computational complexity of keypoint matching algorithms to be as low as possible.
Mohammad Soltanian, Ali Tavakol
core   +1 more source

World Models for Robotic Manipulation: A Survey

SmartBot, EarlyView.
World models enable robots to anticipate how candidate actions may alter object configurations, contact states, and scene geometry before physical execution. This survey examines world models for robotic manipulation as action‐conditioned predictive systems and organizes the literature according to predicted representations, interfaces between ...
Fangyuan Wang, Ziyuan Wang, Guorui Pei, Mengshi Zhang, Canxi Liang, Jun Hu, Zhongxuan Li, Jinsong Wu, Ning Han, Zeqing Zhang, Jiaming Qi, Hongmin Wu, Shiyao Zhang, Pai Zheng, Jia Pan, David Navarro‐Alarcon, Sichao Liu, Peng Zhou   +17 more
wiley   +1 more source

SIFT keypoint descriptors for range image analysis

, 2008
This paper presents work in progress to extend the two-dimensional (2D) Scale Invariant Feature Transform (SIFT) into the 2.5 dimensional (2.5D) domain. Feature descriptors are extracted from range images of human faces and the form of these descriptors ...
Siebert, J., Lo, T-W.R.
core   +3 more sources

A Review on the Development and Key Technologies of Embodied Intelligent Industrial Robots in Smart Manufacturing

SmartBot, EarlyView.
Conventional industrial robots rely on predefined workflows and struggle to adapt to changing tasks and uncertain manufacturing environments. This review examines the development and key technologies of embodied intelligent industrial robots (EIIRs), highlighting a closed‐loop architecture that integrates human and industrial inputs, embodied ...
Jie Wang, Mingsheng Li, Yiming Jiang, Hui Zhang, Zhiqiang Miao, Weixing Peng, He Xie, Yaonan Wang   +7 more
wiley   +1 more source

Multi-Object Keypoint Detection and Pose Estimation for Pigs

Monitoring the daily status of pigs is crucial for enhancing their health and welfare. Pose estimation has emerged as an effective method for tracking pig postures, with keypoint detection and skeleton extraction playing pivotal roles in this process ...
Guo, Qinghua, Pei, Dawei, Sun, Yue, Bijma, Piter, Bolhuis, J. Elizabeth, Martinsen, Kristine Hov, Orsini, Clémence A.E.M., Nordbø, Øyvind, Langenhuizen, Patrick P.J.H.; id_orcid, de With, Peter H.N.   +9 more
core   +1 more source

Trustworthy Imagination in Embodied World Models: Structural Constraints, Reality Anchoring, and Trust Adjudication

SmartBot, EarlyView.
World models let robots imagine alternative futures, such as successfully grasping a glass or dropping it, but visual plausibility alone does not establish trustworthiness. Here, the authors review structural constraint, reality anchoring, and trust adjudication as complementary mechanisms for shaping generation, grounding predictions in real evidence,
Chenwei Shi, Xueyu Luan, Shuo Jiang, Bin He   +3 more
wiley   +1 more source