This is a research group led by Dr Avinash Sharma, who has recently moved to Computer Science & Engineering department at IIT Jodhpur as an associate professor while continuing his affiliation with Centre for Visual Information Technology (CVIT), IIIT Hyderabad. We primarily work on the computer vision and machine learning problems with a focus on 3D scene understanding and reconstruction. Specifically, we deal with problems where human body is reconstructed from 2D images and analysed in 3D, generation & analysis of 3D terrains, registration of point clouds of indoor scenes captured from commodity sensors as well as large outdoor scenes captured from LIDAR scanners.
We are looking for motivated PhD students and long-term RAs/interns with interest in 3D computer vision to join us. Please check here for more details.
Research Interests
- 3D reconstruction of humans
- 3D reconstruction of hands
- Terrain Generation
- 6-DOF pose estimation of rigid objects
- Brain Graph Analytics
- GraphSLAM in robotics
Datasets
|
3DHumans: High-Fidelity 3D Scans of People in Diverse Clothing Styles
|
|
4DHumans: Temporally Consistent Dynamic Scans of Humans in Motion
|
Publications
|
CLIP-Head: Text-Guided Generation of Textured Neural Parametric 3D Head Models
|
|
ConVol-E: Continuous Volumetric Embeddings for Human-Centric Dense Correspondence Estimation
|
|
SHARP: Shape-Aware Reconstruction of People in Loose Clothing
|
|
REF-SHARP: REFined face and geometry reconstruction of people in loose clothing
|
|
Adaptive & Multi-Resolution Procedural Infinite Terrain Generation with Diffusion Models and Perlin Noise
|
|
Discretization-Agnostic Deep Self-Supervised 3D Surface Parameterization
|
|
Supervision by Landmarks: An Enhanced Facial De-occlusion Network for VR-based Applications
|
|
Deep Self-supervised UV Parameterization for Template-free Textured 3D Digitization of Garments
|
|
xCloth: Extracting Template-free Textured 3D Clothes from a Monocular Image
|
|
Robust 3D Garment Digitization from Monocular 2D Images for 3D Virtual Try-On Systems
|
|
Automated tree generation using grammar & particle system
|
|
Coarse-to-fine 3D clothed human reconstruction using peeled semantic segmentation context
|
|
Attention based Occlusion Removal for Hybrid Telepresence Systems
|
|
A-GHN: Attention-based Fusion of Multiple GraphHeat Networks for Structural to Functional Brain Mapping
|
|
GlocalNet: Class-aware Long-term Human Motion Synthesis
|
|
PeeledHuman: Robust Shape Representation for Textured 3D Human Body Reconstruction
|
|
AFN: Attentional Feedback Network based 3D Terrain Super-Resolution
|
|
Feedback Neural Network based Super-resolution of DEM for generating high fidelity features
|
|
Action Quality Assessment using Siamese Network-Based Deep Metric Learning
|
|
DeepHuMS: Deep Human Motion Signature for 3D Skeletal Sequences
|
|
N2NSkip: Learning Highly Sparse Networks using neuron-to-neuron skip connections
|
|
HumanMeshNet: Polygonal Mesh Recovery of Humans
|
|
Resting state dynamics meets anatomical structure: Temporal multiple kernel learning (tMKL) model
|
|
Non-rigid Registration using Spectral Graph Wavelet Features
|
|
SplineNet: B-spline neural network for efficient classification of 3D data
|
|
Towards View-Invariant Intersection Recognition from Videos using Deep Network Ensembles
|
|
Deep Textured 3D Reconstruction of Human Bodies
|
|
Fast multi model motion segmentation on road scenes
|
|
MKL Based Local Label Diffusion for Automatic Image Annotation
|
|
Multiple Kernel Learning Model for Relating Structural and Functional Connectivity in the Brain
|
|
Multi-trajectory pose correspondences using scale-dependent topological analysis of pose-graphs
|
|
SLAM Pose-Graph Robustification Via Multi-Scale Heat-Kernel Analysis
|
|
|
Combining Multiscale Diffusion Kernels for Learning the Structural and Functional Brain Connectivity
|
|
|
Image Annotation using Multi-scale Hypergraph Heat Diffusion Framework
|