Abinash Dutta
I am a Computer Science Engineer who loves building disruptive solutions for problems. I am always eager to learn new things and constantly try to keep on improving. I am a great team player as well as a proficient Leader.
I am a Computer Science Engineer who loves building disruptive solutions for problems. I am always eager to learn new things and constantly try to keep on improving. I am a great team player as well as a proficient Leader.
Full stack developer with experience in MERN, Spring Boot, Java, Python, AWS, Databases, Devops. Worked in state of the art projects across various clients.
Research and development of Machine Translation Tools for low resource language like Assamese. Experience with performance analysis, optimizations and benchmark evaluations for tasks related to Natural Language.
Developed a system which could generate stories based on the contexts given by users under supervision of Prof. Nikhil R. Pal. Incorporated various Technologies like Tensorflow, Keras, Numpy to generate a efficient model. Experience with Position Embedding, Multi Head Attention, Long Short Term Memory Units.
Developed a system for real time hand gesture recognition under the supervision of Dr. Adway Mitra. Incorporated various Technologies like OpenCV, Keras, Numpy to generate a efficient model. Experience with Bounding boxes, Object localization, Object Classification
GPA: 8.83
Percentage: 94.8
Developed a system that in realtime scrapes reviews from the Flipkart site for a product and employs a Fake Review Detection on the reviews. Then based on the true reviews, it does Sentiment Analysis to give result. The Fake Review feature extractor was modeled using Convolutional Neural Networks with 1D convolution and the Sentiment Analysis model was based on a combination of bag of words and n-gram model.
Developed a web toolkit as a part of the final year project which could generate various captions for the images that the user uploads. There was also the provision to generate captions across various machine learning models.
Developed a system to translate English Documents to Hindi keeping the Semantics intact. It was based on a Encoder-Decoder model with Attention which Significantly improved the results.
Developed a Web Interface with a MySQL Database using which a user could search any team, players based on the roles they had. The User Interface was designed using Bootstrap and the backend using NodeJS.
Deployed a system to recognize Hand Gestures in real time. YOLO(You Only Look Once) Algorithm was employed for object localization.
Used a Long Short Term Memory Units to develop a Text Generator. Both Word based as well as Character based models were compared on the Text that were generated.
Apart from being a Software Developer, I enjoy my time being active, be it in the gym or running outdoors. I also enjoy a good cycling sessions along the countryside once in a while.
When forced indoors, I am an aspiring chef and have been cooking since I was 8, and I spend a large amount of my free time exploring the latest technology advancements in the software devlopment and machine learning world.