Astitva Srivastava


Google Ph.D. Fellow | 3D Vision/Graphics Researcher
@ CVIT, IIIT Hyderabad

Qualcomm Innovation Fellowship India'23 Recipient | TCS Research Fellowship'23 Winner

Research Interests: 3D Garment Modelling, Digital Human Avatars, 3D Virtual Try-on



Publications

WordRobe: Text-Guided Generation of Textured 3D Garments
Astitva Srivastava, Pranav Manu, Amit Raj, Varun Jampani, Avinash Sharma | ECCV 2024

CLIP-Head: Text-Guided Generation of Textured Neural Parametric 3D Head Models
Pranav Manu, Astitva Srivastava, Avinash Sharma | SIGGRAPH Asia 2023 Technical Communications

ConVol-E: Continuous Volumetric Embeddings for Human-Centric Dense Correspondence Estimation
Amogh Tiwari, Pranav Manu, Nakul Rathore,Astitva Srivastava, Avinash Sharma | CVPRw 2023

SHARP: Shape-Aware Reconstruction of People in Loose Clothing
Sai Sagar Jinka, Astitva Srivastava, Chandradeep Pokhariya, Avinash Sharma | IJCV 2023

Discretization-Agnostic Deep Self-Supervised 3D Surface Parameterization
Chandradeep Pokhariya, Shanthika Naik, Astitva Srivastava, Avinash Sharma | Sigg. Asia 2022

xCloth: Extracting Template-free Textured 3D Clothes from a Monocular Image
Astitva Srivastava, Chandradeep Pokhariya, Sai Sagar Jinka, Avinash Sharma | ACM Multimedia 2022

Robust 3D Garment Digitization from Monocular 2D Images for 3D Virtual Try-On Systems
Sahib Majithia, Sandeep N. Parameswaran, Sadbhavana Babar, Vikram Garg, Astitva Srivastava, Avinash Sharma | WACV 2022

Experience

Research Intern

Animaker

Worked on the problem of syncing lips of animated 2D characters to a given speech.

Aug '22 - Feb '23

Research Intern

Myntra

Worked on a template-based 3D virtual try-on pipeline and published a paper.

May '21 - Oct '21

Computer Vision Intern

DreamVu

Developed a real-time 3D human performance capture system as a project given by Sony (Japan)

Nov '20 - Mar 2021

Projects

4DHumans
This project aims to capture temporally consistent sequences of humans performing different actions. The data capture system is consist of seven Microsoft Kinect cameras

3DHumans
3DHumans dataset provides 180+ high-fidelity meshes of people in diverse body shapes in various garments styles and sizes. We cover a wide variety of clothing styles, ranging from loose robed clothing, like 'saree' (a typical South-Asian dress) to relatively tight fit clothing, like shirts and trousers.

Peeled Ray Tracer
An extended version of recursive ray tracing for adding realistic lighting mechanism to peelmaps from PeeledHuman representation.

Prompt-SAM
Added CLIP-based text prompting support to Meta's Segment Anything Model (SAM).

3D Point Cloud Segmentation
Implementation of 3D point cloud segmentation using submanifold SparseConv Encoder-Decoder, integrated with Open3D-ML pipeline.

Hand Gesture Recognition
Image processing based mathematical model for detecting hands and recognizing gestures from a custom dataset created from scratch.