Jan 2025 – Jun 2025 · 5 mos · Bangalore Urban district, India · On-site

Jun 2023 – Jul 2023 · 1 mo · Varanasi, Uttar Pradesh, India · On-site

• During my captivating 6-week internship at IIT BHU, I had the incredible opportunity to delve into the dynamic realms of machine learning and deep learning.
• 1) Successfully infused transformer architecture into SegNet, a traditional CNN-based model, to enhance dental image segmentation results and performance metrics.
• 2 )Successfully led architectural modifications to TransSEG and TransUNET models, incorporating additional convolution and deconvolution layers, dilated convolutions, and skip connections, resulting in improved performance and accuracy.
• 3) Pioneered the development of a novel hybrid loss function combining Dice loss and BCE loss, leading to significant enhancements across multiple datasets.
• 4) Experimented with different factor combinations for BCE and Dice Loss to fine-tune the hybrid loss function and further boost model accuracy.
• 5) Introduced additional evaluation metrics, such as Dice coefficient graph and accuracy graph, to comprehensively assess model performance during training and validation.

Sep 2022 – Sep 2024 · 2 yrs · Patna, Bihar, India

• Happy to be a part of IEEE SB NIT Patna.

May 2022 – Aug 2023 · 1 yr 3 mos · Patna, Bihar, India

• I am glad to be a member of Hackslash NITP in team grey Interface which deals with the training of model using machine learning. Hoping to learn a lot!