Upload for open document test api machine learning

test

1. American University of Science and Technology Faculty of Engineering & Computer Science

2. American University of Science and Technology Faculty of Engineering & Computer Science Department of Computer Science Senior Design Proposal Defense – Fall 23-24 23

3. American University of Science and Technology Faculty of Engineering & Computer Science Department of Computer Science Senior Design Proposal Defense – Fall 23-24

4. According to WHO [1], i t is estimated that by 2050 1 in every 10 people will have disabling hearing loss.

5. That’s 10% or 700 million people of the world’s population In Lebanon, this data projection means that around half a million are impacted.

6. In Lebanon, this data projection means that around half a million will be impacted.

7. American University of Science and Technology Faculty of Engineering & Computer Science ن : Notation Deaf and other individuals, such as those unable to physically speak, those who have trouble with oral language due to a disability or condition, face difficulties..

8. At home: Communication Challenges: Lebanese deaf people (LDP) struggle to communicate at home due to family members' limited sign language skills or lack of access to assistive devices.

9. In education: Educational Accessibility: In a recent study done on a sample of LDF, only 7% are bachelor’s degree holders with only 1% holding a doctorate degree.

10. c At the workplace: Discrimination: For the LDP, the Lebanese society did not treat them well: they were stigmatized, shamed, bullied, avoided, neglected and treated with pity in 62% of the cases, while only 38% considered they were well treated.

11. In social life: Lack of sign language interpretes : 60% of the Lebanese Deaf report that such service is unavailable, whereas 40% report its availability with a 70% rate of service satisfaction.

12. Solution Existing solutions for American Sign Language (ASL): Video Relay Services (VRS): Enables sign language communication via video technology. ASL Apps and Online Resources: Offer tutorials and dictionaries for ASL learning. Text-to-Sign Software: Translates written text into ASL signs. Wearable Devices: Translate sign language gestures into text or speech for accessibility.

13. Solution Absence of solutions for Lebanese Sign Language (LSL): Limited References: Insufficient documentation and references hinder the study and preservation of Lebanese Sign Language (LSL). Gesture Databases Needed: The absence of comprehensive databases for LSL gestures restricts research and technological development. Technology based tools: The lack of tailored technological tools for LSL creates barriers to communication and accessibility for the deaf community.

14. Solution Using Technology as a Bridge Bridging Communication Gaps: - Technology connects deaf individuals. - Enhances accessibility and understanding. Importance of LSL Model: - Essential for documenting and preserving LSL. - Facilitates study and cultural preservation.

15. Solution 1) Dataset Creation for LSL: Overview: - Gather video recordings of LSL gestures and conversations. - Transcribe and annotate gestures, expressions, and context. - Ensure diversity and representativeness across speakers and scenarios.

16. Solution 2) Model Development: Introduction to Model Architecture: - Utilizes deep learning algorithms for gesture recognition. - Incorporates convolutional and recurrent neural networks. Importance of Accuracy: - Accurate recognition vital for effective communication. - Ensures accessibility and understanding for deaf individuals.

17. Solution 3) Model Application in Workplaces and Social Life: Application Areas: - Utilized in job interviews, meetings, social gatherings, events, and online platforms.. Benefits: - Improves workplace inclusion by facilitating communication for deaf employees. - Enhances social interactions and participation in diverse social settings. Example: - Integration into platforms like Instagram Stories as an LSL translator showcases its versatility and impact on daily interactions.

18. Conclusions Impact of Technology: - Breaks barriers for the deaf community, enabling equal participation. - Facilitates education, employment, and social interactions. Project Highlights : - Leveraged Tensorflow and Keras to build an LSTM Sign Language Detection Model. - Successfully detects three gestures of three different colors . - A significant step towards advancing accessible communication technology.

19. American University of Science and Technology Faculty of Engineering & Computer Science Let’s watch a demo .

20. According to WHO [1], i t is estimated that by 2050 1 in every 10 people will have disabling hearing loss.

21. Globally, 1.5 billion people (nearly 5% of the population) have hearing loss, with 430 million experiencing disabling hearing loss. With Causations going back to Genetic factors, Infections, Aging, Injuries and many more. We focus on Enabling. Dots or circles to represent the numbers globally, like 5%, then about the 1 in 10 people by 2050. And the mute. Then another page will include those 2 group of dots each colored in a way. All those people require uses sign language to communicate Hello my name is dot. And will be accompanying you throughout all our video. This is your world! Sound and videos of people speaking all over all over (+ a videoshot of Duolingo). Then dot. Tells the screen. This is mine. Complete silence or maybe even white noise. While it’s calming. It’s somehow obscure for my communication. According to WHO I will be 1 in 10 people in 2050 to have difficulties hearing. And alongside our mute friends. We face obstacles at home, at school, at work, at uni. We rely on sign language. But not everybody knows how to use it. (here like the bluberry slide. Jump back and show some people who can actually use sign language, or no). There is a gap. Of communication and of acknowledgment. I’m not just a number. // I’m not the dot in the etc. I am here to bridge. Because i can. I can Navigate, I can inspire, bla bla bla . Ther is a huge stigma in combining deaf with mute

22. Dots or circles to represent the numbers globally, like 5%, then about the 1 in 10 people by 2050. And the mute. Then another page will include those 2 group of dots each colored in a way. All those people require uses sign language to communicate. And very few people can understand the sign language. The slide of the camera should be camera taking a photo illustration, (or 3d model), and then machine learning layers or models that detects those pictures and interpret them into spoken and written output. Hello my name is dot. And will be accompanying you throughout all our video. This is your world! Sound and videos of people speaking all over all over (+ a videoshot of Duolingo). Then dot. Tells the screen. This is mine. Complete silence or maybe even white noise. While it’s calming. It’s somehow obscure for my communication. According to WHO I will be 1 in 10 people in 2050 to have difficulties hearing. And alongside our mute friends. We face obstacles at home, at school, at work, at uni. We rely on sign language. But not everybody knows how to use it. (here like the bluberry slide. Jump back and show some people who can actually use sign language, or no). There is a gap. Of communication and of acknowledgment. I’m not just a number. // I’m not the dot in the etc. I am here to bridge. Because i can. I can Navigate, I can inspire, I can be your hero baby. Ther is a huge stigma in combining deaf with mute

23. ICAN Navigate