Sameer Ambekar

Sameer is working as Research Intern and pursuing Masters in Artificial Intelligence (MSc AI) at the University of Amsterdam (UvA). At UvA, he works under the guidance of Prof. Xiantong Zhen and Prof. Cees Snoek. At UvA, I am working on Test-Time Adaptation of classifiers (🔵->🟦->🔷), Meta learning and Domain Generalization for efficient multi source training and transferrable feautres.

Previously, Sameer worked as Research Assistant at IIT Delhi India under the guidance of Prof. Prathosh A.P. in the field of Domain Adaptation and Analysis of head and neck cancer Hisptopathological Images. Before working at IITD, Sameer worked at Indian Council of Medical Research (ICMR) as a Researcher under the guidance of Dr. Subarna Roy (Scientist G and Director ICMR-NITM), and Mr. Pramod Kumar (Scientist C).

I eagerly look forward to delving deeper into my passion for Deep learning and Computer Vision in my academic endeavors, and actively seeking internship opportunities.

Hobbies: Playing the Flute 🎵, Swimming 🏊, Photography 📷

I am interested in computer vision and deep learning.

Specifically: Domain Generalization, Domain Adaptation, Meta learning, Medical Imaging

We cast the skin segmentation problem as that of target-independent Unsupervised Domain Adaptation (UDA) where we use the data from the Red-channel of the visible-range to develop skin segmentation algorithm on NIR images. We propose a method for target-independent segmentation where the ‘nearest-clone’ of a target image in the source domain is searched and used as a proxy in the segmentation network trained only on the source domain. We prove the existence of ‘nearest-clone’ and propose a method to find it through an optimization algorithm over the latent space of a Deep generative model based on variational inference.

We propose a novel work named SKDCGN that attempts knowledge transfer using Knowledge Distillation (KD). In our proposed architecture, each independent mechanism (shape, texture, background) is represented by a student 'TinyGAN' that learns from the pretrained teacher 'BigGAN'. We demonstrate the efficacy of the proposed method using state-of-the-art datasets such as ImageNet, and MNIST by using KD and appropriate loss functions. Moreover, as an additional contribution, our paper conducts a thorough study on the composition mechanism of the CGNs, to gain a better understanding of how each mechanism influences the classification accuracy of an invariant classifier.

In this academic project at UvA, the objective was to reproduce papers that have been accepted for publication at top conferences

We introduce a state-of-the-art technique, termed as Twin Augmentation, for modifying popular pre-trained deep learning models. Twin Augmentation boosts the performance of a pre-trained deep neural network without requiring re-training. Experiments show, that across a multitude of classifiers, Twin Augmentation is very effective in boosting the performance of given pre-trained model for classification in imbalanced settings.