Saidinesh Pola

I'm Saidinesh Pola, and I have a year of experience working with deep learning methods in computer vision and natural language processing. In my life, I am constantly eager to learn new things. I am competitive when it comes to things that I am passionate about. During the holidays, I enjoy participating in machine learning hackathons and learning about new AI models and techniques. Because of my previous expertise, I can work efficiently on both solo and group projects. Currently, I am working as a front-end developer for the Infosys Interview Platform and, parallelly, I'm working on building an RL agent for Citylearn.

• Secured 3rd Position in AI crowd’s Food Recognition challenge 2022 on Round 1.
• Secured 3rd position in ML-Autism challenge by REVA university and Infosys
• Consistent top 40 ranker in Interviewbit’s coding contests from December 2020 to March 2021

• Food detection algorithm: Trained FAIR’s detectron2 for instance segmentation using MASK R-CNN with various new techniques, which produced third position with 38.1 mAp in AICROWD’s Food Recognition Benchmark 2022 challenge.
• Infosys Interview Platform: Working as a front-end developer for the interview platform, which has so far conducted a million interviews. I am responsible for creating the Extraction Resume parser UI which does the extraction of information from the resume using the BERT transformer.
• Java code generation using GPT-J(6 Billion parameter Language model): Finetuned ethirium-6B model to generate Java code based on the input text. For data generation I have extracted top java GitHub repositories on github and Infosys's github using Google TPU research cloud .
• Low Overhead Sequence Preservation: implementation of a new order-preserving lossless data compression scheme to reduce overhead in joint source and channel coding.
• Deepspeech Automatic Speech Recognition. Finetuned Deepspeech model for automatic translation of Indian English to text for interview platform.

• Food detection algorithm: Trained FAIR’s detectron2 for instance segmentation using MASK R-CNN which produced third position with 38.1 mAp in aicrowd’s Food recognition benchmark 2022 challenge.
• Java code generation using GPT-J(6 Billion parameter Language model): Finetuned ethirium-6B model to generate Java code based on the input text. For data generation I have extracted top java GitHub repositories in github and Infosys's github using google TPU research cloud .
• Low Overhead Sequence Preservation: implementation of a new order-preserving Lossless Data Compression scheme to reduce overhead in joint source and channel coding.

• Food detection algorithm: Trained FAIR’s detectron2 for instance segmentation using MASK R-CNN with various new techniques, which produced third position with 38.1 mAp in AICROWD’s Food Recognition Benchmark 2022 challenge.
• Infosys Interview Platform: Working as a front-end developer for the interview platform, which has so far conducted a million interviews. I am responsible for creating the Extraction Resume parser UI which does the extraction of information from the resume using the BERT transformer.
• Java code generation using GPT-J(6 Billion parameter Language model): Finetuned ethirium-6B model to generate Java code based on the input text. For data generation I have extracted top java GitHub repositories on github and Infosys's github using Google TPU research cloud .
• Low Overhead Sequence Preservation: implementation of a new order-preserving lossless data compression scheme to reduce overhead in joint source and channel coding.
• Deepspeech Automatic Speech Recognition. Finetuned Deepspeech model for automatic translation of Indian English to text for interview platform.