sagniklp Goto Github PK

followers: 59.0 following: 11.0 repos: 33.0 gists: 8.0

Name: Sagnik Das

Type: User

Company: Stony Brook University

Bio: Ph.D. Candidate

Blog: https://www3.cs.stonybrook.edu/~sadas/

Hi there 👋

Hello, I am Sagnik Das.
Currently a PhD Candidate at Stony Brook University Computer Science Department.
My general research interests lie in the intersection of computer vision and computer graphics.

About Me:

🔭 I’m currently working on: Document Unwarping or recovering flatbed scanned images from single RGB image of a document.
🌱 I’m currently learning: Neural rendering/inverse rendering, Illumination modeling (SVBRDF etc.).
👯 I’m looking to collaborate on: Neural Inverse Rendering.
🤔 I’m looking for help with: Non-rigid deformation modeling.
💬 Ask me about: ML, DL, Computer Vision, Computer Graphics, NLP. Happy to help anytime.
📫 How to reach me: PM me on LinkedIn or Email me at sadas[at]cs[dot]stonybrook[dot]edu

Sagnik Das's Projects

db

A PyToch implementation of "Real-time Scene Text Detection with Differentiable Binarization".

This code implements the approach for the following research paper: Fight ill-posedness with ill-posedness: Single-shot variational depth super-resolution from shading; B. Haefner, Y. Quéau, T. Möllenhoff, D. Cremers; Computer Vision and Pattern Recognition (CVPR), 2018.

dewarpnet-demo

Website for DewarpNet demo

dewarpnet-user-study

dewarpnet-webpage

Webpage for DewarpNet paper

disfluency-removal-api

Disfluency Detection, Removal & Correction: Increase Apparent Public Speaking Fluency By Speech Augmentation (ICASSP '19)

doc3d-renderer

Blender rendering codes for doc3D-dataset (Paper: https://www3.cs.stonybrook.edu/~cvl/projects/dewarpnet/storage/paper.pdf)

doyouevenlearn

Essential Guide to keep up with AI/ML/CV/UNameIt

facial-feature-localization

gan

Resources and Implementations of Generative Adversarial Nets: GAN, DCGAN, WGAN, CGAN, InfoGAN

gan_theories

Resources and Implementations of Generative Adversarial Nets which are focusing on how to stabilize training process and generate high quality images: DCGAN, WGAN, EBGAN, BEGAN, etc.