Proposed framework: (a) Training of an SVM classifier using ResNet-50 features, (b) Posture classification using the trained SVM model and ResNet-50 features.

Proposed framework: (a) Training of an SVM classifier using ResNet-50 features, (b) Posture classification using the trained SVM model and ResNet-50 features.

<p>Proposed framework: (a) Training of an SVM classifier using ResNet-50 features, (b) Posture classification using the trained SVM model and ResNet-50 features.</p>

Saved in:

Bibliographic Details
Main Author:	Irfan Kareem (20390414) (author)
Other Authors:	Syed Farooq Ali (20390417) (author), Muhammad Bilal (737265) (author), Muhammad Shehzad Hanif (18980477) (author)
Published:	2024
Subjects:	Biotechnology Space Science Biological Sciences not elsewhere classified Information Systems not elsewhere classified using four postures taking another step support vector machine lighting conditions due existing methods yielding shown promising results proposed approach outperforms human posture recognition ur fall detection based hybrid architecture 9 %, 97 82 %, 97 deep residual network fall detection performance fall detection residual network results demonstrate posture task camera fall 33 %, network models deep learning svm outperforms xlink "> time efficiency three datasets realistic environment random forest paper analyzes naive bayes mlp used last decade decision tree comparison purposes classifier layer classification layer also compared achieved 100 95 %,
Tags:	Add Tag No Tags, Be the first to tag this record!

Cannot write session to /tmp/vufind_sessions/sess_eghngg9abo932u7im6rmfbv44n