Citation :

Seema Bhalgaonkar, Mousami Munot, Alwin Anuse, "Distribution-Aware Dynamic Temperature Knowledge Distillation for Efficient Indian Sign Language Recognition on Edge Devices," International Journal of Engineering Trends and Technology (IJETT), vol. 74, no. 10, pp. 362-377, 2026. Crossref, https://doi.org/10.14445/22315381/IJETT-V74I10P126

Abstract

Sign language recognition serves as a critical communication aid for the deaf-mute community, enabling seamless interaction with the hearing population. While deep learning models have demonstrated remarkable recognition accuracy, their large model size makes deployment on resource-constrained edge devices challenging. Knowledge distillation is highly effective in compressing heavy models; however, the classic knowledge distillation approach relies on a static temperature setting, which limits the student model's ability to acquire rich inter-class knowledge from the teacher. This study validates the effectiveness of the Distribution-Aware Dynamic Temperature Knowledge Distillation framework for Indian Sign Language gesture recognition. MobileNetV2 is employed as the teacher model, and two student configurations - MobileNetV2 α=0.5 and α=0.35, representing moderate and aggressive compression levels respectively, are evaluated across multiple performance parameters. Experimental results demonstrate a consistent accuracy improvement of approximately 6% (α=0.5) and 5% (α=0.35) over classic knowledge distillation across both student configurations, with model size reduced from 33.93MB to 5.76MB and 4.64MB, respectively. The quantized student models achieve a maximum compression ratio of 12.85×, retaining 95.92% recognition accuracy with an inference time of 3.63ms, enabling real-time gesture recognition on Raspberry Pi. These results demonstrate the effectiveness of DAD-DTKD for sign language recognition and its potential for compressed model deployment on edge devices.

Keywords

Edge AI, Dynamic temperature setting, Indian Sign Language, Model compression, Knowledge distillation.

References

1. World Health Organization, Deafness and Hearing Loss, WHO Fact Sheet, 2026. [Online]. Available: https://www.who.int/news-room/fact-sheets/detail/deafness-and-hearing-loss
2. Rina Damdoo, and Praveen Kumar, “An Integrative Survey on Indian Sign Language Recognition and Translation,” IET Image Processing, vol. 19, no. 1, pp. 1-29, 2025.
   [CrossRef] [Google Scholar] [Publisher Link]
3. Aamir Wali et al., “Recent Progress in Sign Language Recognition: A Review,” Machine Vision and Applications, vol. 34, no. 6, 2023.
   [CrossRef] [Google Scholar] [Publisher Link]
4. Karly Kudrinko et al., “Wearable Sensor-based Sign Language Recognition: A Comprehensive Review,” IEEE Reviews in Biomedical Engineering, vol. 14, pp. 82-97, 2021.
   [CrossRef] [Google Scholar] [Publisher Link]
5. Ang Ji et al., “Dataglove for Sign Language Recognition of People with Hearing and Speech Impairment Via Wearable Inertial Sensors,” Sensors, vol. 23, no. 15, pp. 1-20, 2023.
   [CrossRef] [Google Scholar] [Publisher Link]
6. Shagun Katoch, Varsha Singh, and Uma Shanker Tiwary, “Indian Sign Language Recognition System using SURF with SVM and CNN,” Array, vol. 14, pp. 1-9, 2022.
   [CrossRef] [Google Scholar] [Publisher Link]
7. Tangfei Tao et al., “Sign Language Recognition: A Comprehensive Review of Traditional and Deep Learning Approaches, Datasets, and Challenges,” IEEE Access, vol. 12, pp. 75034-75060, 2024.
   [CrossRef] [Google Scholar] [Publisher Link]
8. Karen Simonyan, and Andrew Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” arXiv preprint arXiv, pp. 1-14, 2015.
   [CrossRef] [Google Scholar] [Publisher Link]
9. Kaiming He et al., “Deep Residual Learning for Image Recognition,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770-778, 2016.
   [Google Scholar] [Publisher Link] 
10. Christian Szegedy et al., “Going Deeper with Convolutions,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1-9, 2015.
    [Google Scholar] [Publisher Link]
11. Venus Singla, Seema Bawa, and Jasmeet Singh, “Enhancing Indian Sign Language Recognition through Data Augmentation and Visual Transformer,” Neural Computing and Applications, vol. 36, no. 24, pp. 15103-15116, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
12. M.K. Vidhyalakshmi et al., “Indian Sign Language Recognition using Transfer Learning with Efficient Net,” Smart Science, vol. 12, no. 2, pp. 269-280, 2024.
    [CrossRef] [Google Scholar] [Publisher Link] 
13. Soumen Das, Saroj Kr. Biswas, and Biswajit Purkayastha, “Occlusion Robust Sign Language Recognition System for Indian Sign Language Using CNN and Pose Features,” Multimedia Tools and Applications, vol. 83, no. 36, pp. 84141-84160, 2024.
    [CrossRef] [Google Scholar] [Publisher Link] 
14. Sakshi Sharma, and Sukhwinder Singh, “ISL Recognition System using Integrated Mobile-Net and Transfer Learning Method,” Expert Systems with Applications, vol. 221, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
15. Andrew G. Howard et al., “MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications,” arXiv preprint arXiv, pp. 1-9, 2017.
    [CrossRef] [Google Scholar] [Publisher Link]
16. Mark Sandler et al., “MobileNetV2: Inverted Residuals and Linear Bottlenecks,” 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, pp. 4510-4520, 2018.
    [CrossRef] [Google Scholar] [Publisher Link]
17. Andrew Howard et al., “Searching for MobileNetV3,” Proceedings of the IEEE International Conference on Computer Vision, pp. 1314-1324, 2019.
    [Google Scholar] [Publisher Link]
18. Shilpa Ingoley, and Jagdish Bakal, “Efficient Models based on Deep Learning Technique for Indian Sign Language,” Procedia Computer Science, vol. 258, pp. 1980-1992, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
19. Defu Liu et al., “A Survey of Model Compression Techniques: Past, Present, and Future,” Frontiers in Robotics and AI, vol. 12, pp. 1-16, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
20. Ali Alqahtani, Xianghua Xie, and Mark W. Jones, “Literature Review of Deep Network Compression” Informatics, vol. 8, no. 4, pp. 1-12, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
21. Kequan Lin, Zhenjie Lin, and Shiran Wang, “Model Compression for IoT Edge Devices based on Knowledge Distillation,” International Journal of Pattern Recognition and Artificial Intelligence, pp. 1-21, 2026.
    [CrossRef] [Google Scholar] [Publisher Link]
22. Yang He, and Lingao Xiao, “Structured Pruning for Deep Convolutional Neural Networks: A Survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 5, pp. 2900-2919, 2024.
    [CrossRef] [Google Scholar] [Publisher Link] 
23. Hongrong Cheng, Miao Zhang, and Javen Qinfeng Shi, “A Survey on Deep Neural Network Pruning: Taxonomy, Comparison, Analysis, and Recommendations,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 46, no. 12, pp. 10558-10578, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
24. Amir Gholami et al., A Survey of Quantization Methods for Efficient Neural Network Inference, 1^st ed., Low-Power Computer Vision, Chapman and Hall/CRC, pp. 1-33, 2022.
    [Google Scholar] [Publisher Link]
25. Kalyan Cherukuri, and Aarav Lala, “Low-Rank Matrix Approximation for Neural Network Compression,” arXiv preprint arXiv, pp. 1-5, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
26. Jianping Gou et al., “Knowledge Distillation: A Survey,” International Journal of Computer Vision, vol. 129, no. 6, pp. 1789-1819, 2021.
    [CrossRef] [Google Scholar] [Publisher Link]
27. Yukang Wei, and Yu Bai, “Dynamic Temperature Knowledge Distillation,” arXiv preprint arXiv, pp. 1-15, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
28. Zinah Raad Saeed et al., “A Systematic Review on Systems-based Sensory Gloves for Sign Language Pattern Recognition: An Update from 2017 to 2022,” IEEE Access, vol. 10, pp. 123358-123377, 2022.
    [CrossRef] [Google Scholar] [Publisher Link]
29. Deepti Aggarwal et al., “Sign Language Prediction using Machine Learning Techniques: A Review,” 2023 Second International Conference on Electronics and Renewable Systems, Tuticorin, India, pp. 1296-1300, 2023.
    [CrossRef] [Google Scholar] [Publisher Link]
30. Sibgat Ul Islam et al., “Dynamic Temperature Scheduler for Knowledge Distillation,” 2025 IEEE International Conference on Big Data (BigData), Macau, China, pp. 1412-1417, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
31. Yu-e Lin et al., “ATMKD: Adaptive Temperature Guided Multi-Teacher Knowledge Distillation,” Multimedia Systems, vol. 30, no. 5, 2024.
    [CrossRef] [Google Scholar] [Publisher Link]
32. Wenjie Liu et al., “Be your Own Doctor: Temperature Scaling Self-Knowledge Distillation for Medical Image Classification,” Neurocomputing, vol. 638, pp. 1-12, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
33. Ismail Lamaakal et al., “An Explainable Hybrid CNN-Transformer Model for Sign Language Recognition on Edge Devices using Adaptive Fusion and Knowledge Distillation,” Scientific Reports, vol. 16, no. 1, pp. 1-23, 2026.
    [CrossRef] [Google Scholar] [Publisher Link]
34. Shunzhi Yang et al., “Adaptive Temperature Distillation Method for Mining Hard Samples’ Knowledge,” Neurocomputing, vol. 636, 2025.
    [CrossRef] [Google Scholar] [Publisher Link]
35. Geoffrey Hinton, Oriol Vinyals, and Jeff Dean, “Distilling the Knowledge in a Neural Network,” arXiv preprint, pp. 1-9, 2015.
    [CrossRef] [Google Scholar] [Publisher Link]
36. Atharva Dumbre, Indian Sign Language (ISLRTC Referred), Kaggle Dataset, 2023. [Online]. Available: https://www.kaggle.com/datasets/atharvadumbre/indian-sign-language-islrtc-referred