Hello there!
My name is Aryan, and I'm an applied scientist working in the research-reasoning team at Amazon, based in Santa Clara, California. Before joining Amazon, I was a Senior Computer Vision Algorithm Engineer at Ambarella.
In the past, I've also had the opportunity to intern at companies like SIEMENS
and EigenControl, where I worked on various machine learning projects.
In 2020, I completed my Ph.D. from the University of Houston, where I worked as a research assistant at the HULA Lab and was a fellow at the Center for Advanced Computing and Data Science (CACDS). My research primarily focused on Machine Learning and its applications in Computer Vision, particularly when dealing with less labeled and imperfect data.
I'm really passionate about exploring different inference methods, neural network architectures (including attention and memory models), and anything that can be described as "machine reasoning." Additionally, I've been involved in several open-source and cool side projects, which you can find at the bottom of this page. I'm always on the lookout for exciting research and professional opportunities, so please feel free to reach out to me. If you'd like to learn more about my research topics and other experiences, you can check out my Resume.
I'm excited to connect with you all and look forward to engaging discussions and collaborations!
Designed, and implemented cutting-edge multi-modal (time-series and video) deep neural network models to enable accurate customer tracking, item identification, and real-time receipt generation with high precision for Amazon's Just Walk Out technology in Amazon Go retail stores.
Designed and implemented advanced computer vision algorithms critical for autonomous driving capabilities, including multi-modal multi-object tracking, segmentation, and detection techniques that enable reliable perception and understanding of the vehicle's surroundings.
Worked with Dr. Ali Kamen and Dr. Tomasso Mansi to design and develop a learned non-rigid point cloud-based deep neural network model for accurate MRI-Ultrasound registration of the prostate, enabling precise multimodal image alignment and enhanced diagnostic capabilities.
Working with Dr. Hien Van Nguyen on representation learning with less label and imperfect data