Unexpressed emotions will die no more!
We at ExpressAbled have an aim of connecting the people with auditory and visual disabilities with the world by implementing various methods to overcome the obstacles of communication.
We at ExpressAbled have an aim of connecting the people with auditory and visual disabilities with the world by implementing various methods to overcome the obstacles of communication.
Converts ASL gestures into text. Helps people with auditory disabilities to communicate fluently with people who do not understand sign language.
Describes the image provided by the user into text caption. Helps people with visual impairment to by generating a textual description of the image.
Captures the voice of the user via microphone and converts it to text. Generates text in real-time with high accuracy just through microphone.
Our website is easy to use with well-defined navigation through all the features. No requirement of any equipment for access which makes it cost-effective.
We have incorporated 3 autonomous features under one website. We have taken care for precision, accuracy and providing instant output in real-time for every feature, keeping in mind the user-friendly interface for easy access.
Live captured images are converted into vectors using pre-trained MobileNet CNN model. This trained dataset is used to predict the new gesture which is done using kNN classification.
Images can be provided by users and can be interpreted by using both Convolutional and recurrent neural networks (Resnet50 architecture and LSTM) to give apt captions for the same.
The user speaks through a microphone, the device detects the voice and converts it to text using open-source Web Speech API to convert speech to text in real-time.
No, there is no requirement of any additional equipment. You will only require a decent PC/laptop with web camera.
No, there are no specifications on image resolution or dimension, just upload the image in jpg or jpeg or png format.
Speech-to-Text feature uses Web-Speech-API which is not compatible with all browsers. We recommend you to use Google Chrome. For more information, please visit this website.
No, you will need to train your model only when you are using it for the first time. Once created, you can download your dataset. The next time you want to use the same dataset, you can simply upload that .json file and you are good to go!
Sardar Patel Institute of Technology, Andheri (W), Mumbai
expressabled@gmail.com
+91 9172494383