General Information:
|
| Year Of Paper Submission : 2021-22 |
| Type of Applicant : Student |
| Selected Course : UG |
| Department of Applicant : IT |
| Class of Applicant : B.E. |
Applicants Details:
|
| UCID of Applicant : 2019140030 |
| Applicant Name : Jai Prasanna Joshi |
| Applicant Email : jai.joshi@spit.ac.in |
| Applicant Contact Number : 9987273240 |
| UCID of Applicant No. 2 : 2019140020 |
| Applicant No. 2 Name : Parshav Gandhi |
| Applicant No. 2 Email : parshav.gandhi@spit.ac.in |
| Applicant No. 2 Contact Number : 9920631031 |
Guide Details:
|
| Department of Guide No. 1 : Information Technology |
| Name of First Guide : Prof. Rupali Sawant |
Paper Details:
|
| Title of Paper : Sign Language Certification Platform with Action Recognition using LSTM Neural Networks |
| Type of Paper : international |
| Type of Publication : international |
| Name of the Conference/journal/publisher :
IEEE Xplore |
| Date Of Conference / Journal / Book :
2022-06-23 |
| Conference_Type : ieee |
| Name of the Hosting Institute of the Conference :
SCMS School of Engineering and Technology |
| Address of Host Institute : Vidya Nagar, Palissery, Palissery, Karukutty, Kerala 683576 |
| Remark : The American Sign Language substantially facilitates communication among people that are either hard of hearing or mute. According to a few research studies, at least 70 million people all across the world communicate through sign language. However, there are only a few hundred thousand speakers, limiting the number of persons with whom they can interact comfortably. Alternative modes of communication, such as written communication, can be inconvenient, impersonal, and even cumbersome on a daily basis, and even more so in an emergency. [5]. We present an ASL learning platform that employs LSTM neural networks to recognize the user‧s gesture (action recognition) and deliver real-time feedback in order to overcome this barrier and enable dynamic communication. According to the experimental results, the accuracy achieved while training ASL words was 99.43%, while training ASL alphabets was 91.01% and while training ASL numbers was 98.80%. |
