Deep Learning Model Fusing Vision and Skeletons Classifies Hockey Actions Accurately
Researchers developed an integrated neural network combining visual features and skeletal graph convolutions that achieved 98.73 percent accuracy in recognizing ...
Researchers developed an integrated neural network combining visual features and skeletal graph convolutions that achieved 98.73 percent accuracy in recognizing ...
Researchers developed a classroom teaching analysis system combining skeletal pose estimation and video appearance features to evaluate teacher and student ...
A multimodal deep learning system combining pose estimation, recurrent neural networks, and a retrieval-augmented chatbot achieved high accuracy in identifying ...
Researchers evaluated ten 3D LiDAR SLAM algorithms in simulated and real-world construction environments, finding tightly coupled inertial systems significantly reduce ...
Agricultural engineers developed SSD-YOLOv11 to simultaneously classify Kushui rose maturity stages and estimate mature bud harvesting orientation under challenging field ...
Researchers developed a multimodal deep-learning framework that evaluates competitive aerobics movements by combining synchronized video, skeletal tracking, wearable inertial data, ...
A leakage-controlled evaluation on the AQA-7 benchmark demonstrates that transparent pose-derived image-plane geometry fails to reliably predict expert sports performance ...
Researchers developed a computer vision and pose estimation framework combining YOLOv8s localization and 17-keypoint skeleton tracking to jointly predict jianzi ...
Researchers evaluate cross-view generalization in pose-based fall detection, demonstrating that sequential Transformers and coordinate augmentations enhance robustness across unseen camera ...
Researchers developed a multi-stream graph convolutional network and dataset to classify 13 martial arts categories on edge hardware, demonstrating development ...
A scoping review of 14 studies finds that vision-based markerless tool tracking in open surgery remains dominated by simulated data ...
Researchers combined YOLOv5, DeepSORT, OpenPose, BiLSTM and Dynamic Time Warping into a long jump teaching system that scored take-off movements ...
A 2026 hospital computer-vision study combined small-object detection, wrist-position checks and memory-enhanced motion analysis to identify hazards and abnormal behavior ...
A lightweight pose-estimation model recognized six fitness actions and counted four types of repetitions with 94.37% accuracy in a 200-video ...
A new study of 500 elite throwing clips shows that modeling the four biomechanical phases of a throw reduces prediction ...
A vision-based machine learning study of 61 dementia care residents found that ceiling-mounted cameras and gait analysis could detect gait ...
The distinction between assessment and control is central to understanding why this framework represents a meaningful step forward for collaborative ...
A new deep learning system distills a high-accuracy pose estimation model into a lightweight edge-deployable AI coach that corrects athletes' ...
Researchers released QAMQOR, a public benchmark from 47 hours of robot-assisted autism therapy video showing that evaluation protocols, not algorithms, ...
Researchers at Kocaeli University combined RGB video features and skeleton-based motion features in a multimodal deep learning framework that detects ...
We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.
Bioengineer.org © Copyright 2023 All Rights Reserved.
Bioengineer.org © Copyright 2023 All Rights Reserved.