Amogh Dumbre

At Department of Artificial Intelligence and Data Science, VIT Pune

• Developed Medi-Vision, an integrated vision and language model utilizing Vision Transformers (ViT) for medical Visual Question Answering (VQA) tasks, improving diagnostic accuracy by 22%. • Conducted advanced research and implemented techniques such as fine-tuning ViT models and transfer learning, improving model performance by 18%. • Collaborated with healthcare professionals to validate the model’s effectiveness, resulting in a 20% reduction in manual image analysis time.