Affiliation:
1. Raja Rajeswari College of Engineering, Bengaluru, Karnataka, India
Abstract
“TensorLip presents a pioneering approach towards the realm of speech-recognition and communication accessibility through the fusion of deep-learning and TensorFlow technology. Our paper focuses on the advancement of a lip-reading system capable of deciphering spoken language solely from visual cues of lip movements. Leveraging the power of algorithms in deep learning, particularly tailored and optimized within the TensorFlow framework, TensorLip aims to bridge the communication gap in situations where individuals experience hearing challenges or amidst noisy surroundings where traditional audio-based methods fall short. By harnessing the vast potential of neural networks, our innovative solution promises to revolutionize the manner in which we perceive and understand spoken language, thereby enhancing inclusivity and facilitating seamless communication across diverse linguistic and auditory landscapes.”.
Reference14 articles.
1. [1] Enhancing Lip Reading Efficiency with Efficient-GhostNet by Gaoyan Zhang and Yuanyao Lu.
2. [2] Advancing Lip Reading: WLAS Network for Open-World Recognition by Joon Son Chung and Andrew Senior, and Oriol Vinyals, Andrew Zisserman.
3. [3] Sharma, Annu, Praveena Chaturvedi, and Shwetank Arya. "Human recognition methods based on biometric technologies." International Journal of Computer Applications 120.17 (2015).
4. [4] Evolution of Automatic Lip-Reading: A Deep-Learning Perspective by Adriana Fernandez-Lopez, Federico M. Sukno.
5. [5] Advancements in Lipreading: From Traditional to Deep-Learning Methods by Mingfeng Hao, MutallipMamut, NurbiyaYadikar, Alimjan Aysa.