Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I’m an ML Engineer building computer vision and generative AI systems, from GAN/diffusion-based synthetic data generation to multimodal fine-tuning and model compression for edge inference. I focus on improving model robustness under low data, class imbalance, and consistency constraints. I’ve also researched face image quality assessment and medical imaging, delivered practical vision solutions like segmentation/subject masking and exposure correction, and develop end-to-end deep learning pipelines for real-world products.…I’m an ML Engineer building computer vision and generative AI systems, from GAN/diffusion-based synthetic data generation to multimodal fine-tuning and model compression for edge inference. I focus on improving model robustness under low data, class imbalance, and consistency constraints. I’ve also researched face image quality assessment and medical imaging, delivered practical vision solutions like segmentation/subject masking and exposure correction, and develop end-to-end deep learning pipelines for real-world products.

AMAN AGARWAL

AI Engineer, Data Scientist, Programmer, +2





I’m an ML Engineer building computer vision and generative AI systems, from GAN/diffusion-based synthetic data generation to multimodal fine-tuning and model compression for edge inference. I focus on improving model robustness under low data, class imbalance, and consistency constraints. I’ve also researched face image quality assessment and medical imaging, delivered practical vision solutions like segmentation/subject masking and exposure correction, and develop end-to-end deep learning pipelines for real-world products.…I’m an ML Engineer building computer vision and generative AI systems, from GAN/diffusion-based synthetic data generation to multimodal fine-tuning and model compression for edge inference. I focus on improving model robustness under low data, class imbalance, and consistency constraints. I’ve also researched face image quality assessment and medical imaging, delivered practical vision solutions like segmentation/subject masking and exposure correction, and develop end-to-end deep learning pipelines for real-world products.

Available to hire

I’m an ML Engineer building computer vision and generative AI systems, from GAN/diffusion-based synthetic data generation to multimodal fine-tuning and model compression for edge inference. I focus on improving model robustness under low data, class imbalance, and consistency constraints.

I’ve also researched face image quality assessment and medical imaging, delivered practical vision solutions like segmentation/subject masking and exposure correction, and develop end-to-end deep learning pipelines for real-world products.

Skills

Language

Work Experience

Generative AI Research Intern at MURVEN INFOTECH DESIGN SOLUTIONS LLP

January 1, 2023 - March 31, 2023

Developed a mood-based Text-to-Music system using Python and Flask, leveraging NLP for personalized music generation.

Machine Learning Engineer at AFTERSHOOT PVT. LTD.

December 1, 2022 - Present

Worked on image retouching solutions for portrait photography, implementing and training GANs and diffusion models for synthetic data generation. Finetuned multimodal models using LoRA and researched model size reduction for edge inference. Developed a face blur classification approach using vision transformers under low-data, class-imbalance, and consistency challenges; using different loss functions to improve results by 26% on defined metrics. Built subject masking and segmentation techniques improving segmentation accuracy by 15%, and applied quantization to reduce model size by 50%. Also worked on scene recognition and image exposure correction, improving image edits by 13% and reducing user churn by 7%.

Freelancer (Deep Learning) at FIVERR

September 1, 2022 - November 30, 2023

Collaborated on 20+ projects for 15+ clients, building deep learning systems for object detection, segmentation, time-series forecasting, and natural language processing. Delivered end-to-end robust deep learning pipelines from development to deployment considerations.