0% Complete
Home
/
13th International Conference on Computer and Knowledge Engineering
Distilling Knowledge from CNN-Transformer Models for Enhanced Human Action Recognition
Authors :
Hamid Ahmadabadi
1
Omid Nejati Manzari
2
Ahmad Ayatollahi
3
1- School of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
2- School of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
3- School of Electrical Engineering, Iran University of Science and Technology, Tehran, Iran
Keywords :
Action Recognition،Still Image،Deep Learning،Knowledge Distillation
Abstract :
This paper presents a study on improving human action recognition through the utilization of knowledge distillation, and the combination of CNN and ViT models. The research aims to enhance the performance and efficiency of smaller student models by transferring knowledge from larger teacher models. The proposed method employs a Transformer vision network as the student model, while a convolutional network serves as the teacher model. The teacher model extracts local image features, whereas the student model focuses on global features using an attention mechanism. The Vision Transformer (ViT) architecture is introduced as a robust framework for capturing global dependencies in images. Additionally, advanced variants of ViT, namely PVT, Convit, MVIT, Swin Transformer, and Twins, are discussed, highlighting their contributions to computer vision tasks. The ConvNeXt model is introduced as a teacher model, known for its efficiency and effectiveness in computer vision. The paper presents performance results for human action recognition on the Stanford 40 dataset, comparing the accuracy and mAP of student models trained with and without knowledge distillation. The findings illustrate that the suggested approach significantly improves the accuracy and mAP when compared to training networks under regular settings. These findings emphasize the potential of combining local and global features in action recognition tasks.
Papers List
List of archived papers
Cross-project Defect Prediction with An Enhanced Transfer Boosting Algorithm
Nazgol Nikravesh - Mohammad Reza Keyvanpour
Leveraging Self-Supervised Models for Automatic Whispered Speech Recognition
Aref Farhadipour - Homa Asadi - Volker Dellwo
Non-Negative Matrix Factorization improves Residual Neural Networks
Hojjat Moayed
An optimal workflow scheduling method in cloud-fog computing using three-objective Harris-Hawks algorithm
Ahmadreza Montazerolghaem - Maryam Khosravi - Fatemeh Rezaee
FAST: FPGA Acceleration of Neural Networks Training
Alireza Borhani - Mohammad Hossein Goharinejad - Hamid Reza Zarandi
Automating Theory of Mind Assessment with a LLaMA-3-Powered Chatbot: Enhancing Faux Pas Detection in Autism
Avisa Fallah - Ali Keramati - Mohammad Ali Nazari - Fatemeh Sadat Mirfazeli
Impossible differential and zero-correlatin linear cryptanalysis of Marx, Marx2, Chaskey andSpeck32
Mahshid Saberi - Nasour Bagheri - Sadegh Sadeghi
Speech Emotion Recognition Using a Hierarchical Adaptive Weighted Multi-Layer Sparse Auto-Encoder Extreme Learning Machine with New Weighting and Spectral/SpectroTemporal Gabor Filter Bank Features
Fatemeh Daneshfar - Seyed Jahanshah Kabudian
Towards Study of Research Topics Evolution in Artificial Intelligence based on Topic Embedding
Seyyed Reza Taher Harikandeh - Sadegh Aliakbary - Soroush Taheri
Deep Learning Feature Extraction for COVID-19 Detection Algorithm using Computerized Tomography Scan
Maisarah Mohd Sufian - Ervin Gubin Moung - Chong Joon Hou - Ali Farzamnia
more
Samin Hamayesh - Version 41.3.1