Akshay Gadi Patil

About Me

I am an Applied Scientist at Amazon Science in Sunnyvale, CA, working on GenAI for 3D.

I earned a Ph.D. degree from Simon Fraser University, adviced by Hao.

I work in the areas of 3D computer graphics and vision, developing computational models for generative and interactive shape/scene understanding and geometric modeling problems.

Research Interests : Neural modeling of 3D shapes, 3D scenes and 2D documents, exploring different representations, single and multi-modal, centered around generation, reconstruction, and interaction tasks.

Earlier, I graduated with a M.Tech in Electrical Engineering (Signal Processing) from IIT Gandhinagar, working with Shanmuganathan Raman. I was also a visiting student at JAIST, Japan .

Publications [Scholar]

"Si falta pasión no se encuentra la victoria".

- Rafa

2024

Active Coarse-to-Fine Segmentation of Moveable Parts from Real Images
Ruiqi Wang, Akshay Gadi Patil, Fenggen Yu, Hao Zhang
ECCV 2024 [Paper]

2023

	Learning to Model 3D Indoor Scenes and Articulated Objects Akshay Gadi Patil Ph.D. Dissertation (Distinction: Defence and Thesis passed as is) Link
	Advances in Data-Driven Analysis and Synthesis of 3D Indoor Scenes Akshay Gadi Patil, Supriya Gadi Patil, Manyi Li, Matthew Fisher, Manolis Savva, Hao Zhang Computer Graphics Forum (CGF) 2023 [Paper]
	RoSI: Recovering 3D Shape Interiors from Few Articulation Images Akshay Gadi Patil, Yiming Qian, Shan Yang, Brian Jackson, Eric Bennett, Hao Zhang arXiv 2023 [Paper]
	DiViNeT: 3D Reconstruction from Disparate Views via Neural Template Regularization Aiya Vora, Akshay Gadi Patil, Hao Zhang NeurIPS 2023 [Paper]
	Coarse-to-Fine Active Segmentation of Interactable Parts in Real Scene Images Ruiqi Wang, Akshay Gadi Patil, Fenggen Yu, Hao Zhang arXiv 2023 [ Paper]

2021

LayoutGMN: Neural Graph Matching for Structural Layout Similarity
Akshay Gadi Patil, Manyi Li, Matthew Fisher, Manolis Savva, Hao Zhang
CVPR 2021

Paper, Supp, Code, Video

2020

	DR-KFS: A Differentiable Visual Similarity Metric for 3D Shape Reconstruction Jiongchao Jin, Akshay Gadi Patil, Zhang Xiong, Hao Zhang ECCV 2020 Paper
	READ: Recursive Autoencoders for Document Layout Generation Akshay Gadi Patil, Omri Ben-Eliezer, Or Perel, Hadar Averbuch-Elor CVPR 2020 (Workshop on Text and Documents in Deep Learning Era, *Best Paper Award* ) Paper, Supplementary, Video

2018

	GRAINS: Generative Recursive Autoencoders for INdoor Scenes Manyi Li, Akshay Gadi Patil, Kai Xu, Siddhartha Chaudhuri, Owais Khan, Ariel Shamir, Changhe Tu, Baoquan Chen, Daniel Cohen-Or, Hao Zhang Transactions on Graphics (Special Issue of SIGGRAPH 2019) Project Page
	Language-Driven Synthesis of 3D Scenes Using Scene Databases Akshay Gadi Patil, Rui Ma, Matthew Fisher, Manyi Li, Sören Pirk, Binh-Son Hua, Sai-Kit Yeung, Xin Tong, Leonidas Guibas, Hao Zhang SIGGRAPH Asia, Tokyo, 2018 (*Co-first Authors) Project Page, Media

2016

	Automatic Content-Aware Non-Photorealistic Rendering of Images Akshay Gadi Patil, Shanmuganathan Raman International Symposium on Visual Computing (ISVC), Las Vegas, 2016. Paper
	Tone Mapping HDR Images Using Local Texture and Brightness Measures Akshay Gadi Patil, Shanmuganathan Raman International Conference on Computer Vision and Image Processing (CVIP), 2016.

Awards

Outstanding Reviewer Award, ECCV 2022

Outstanding Reviewer Award, ICCV 2021

Best Paper Award at the CVPR Workshop on Texts and Documents in Deep Learning Era - 2020

SFU Computing Science Graduate Fellowship - 2016, 2018, 2019

SFU Graduate Fellowhsip - 2018

Century 21 Charlwood Family Graduate Scholarship - 2018, 2019

Helmut & Hugo Eppich Family Graduate Scholarship - 2021

Invited Talks and Organized Workshops

Stanford Vision and Learning Lab - May 2023

Max Planck Institute for Informatics - March 2023

Facebook - May 2022

Computational Design and Fabrication Group, MIT - July 2021

Young Scientists' Seminar, CMSA, Harvard University - Dec 2020

Graphics and Vision Seminar, Tel Aviv University - Dec 2018

Amazon - Dec 2018

Co-organizer of Learning to Generate 3D Shapes and Scenes workshop at ECCV 2022