The Vision & Image Understanding (VIU) Lab is a part of the Electrical and Computer Engineering department in Johns Hopkins University. We focus on several theoretical and application aspects of computer vision and image understanding. 

Advancing research in computer vision is one of the most important aspects in developing robust artificial intelligent systems. We believe it is critical to develop strong reasoning abilities about the visual world and in that attempt, we work on several areas related to computer vision and image understanding such as recognition, detection, image restoration, user authentication, crowd analytics, cross spectrum face synthesis, domain adaptation, open set recognition, bio-medical image analysis, biometrics, etc. Specifically, we focus on developing a variety of novel machine learning techniques, such as end-to-end deep learning and neural networks for these applications.

User Active Authentication

We introduce Extremal Openset Rejection (EOR), a two fold mechanism with a sparse representation-based identification step and a verification step for this purpose...

read more

Image-based Crowd Analytics

The study of human behavior based on computer vision techniques has gained a lot of interest in recent years. In particular, the behavioral analysis of crowded scenes ...

read more

One Class Classification

One class classification refers to the problem of identifying decision boundary around a target class. Due to unavailability of any negative data samples during training ...

read more

Multimodal Subspace Clustering

We present convolutional neural network (CNN) based approaches for unsupervised multimodal subspace clustering. The proposed framework consists of ...

read more

Image Restoration (Deraining/Dehazing)

In many applications such as drone-based video surveillance, self driving cars and recognition under night-time and low-light conditions, the captured images and ...

read more

Bio-medical Image Analysis

The main goal of medical image analysis is to extract clinically relevant information or knowledge from medical images. While closely related to the field of medical imaging ...

read more

Cross Spectrum Face Synthesis

Cross-spectrum face recognition refers to the problem of matching faces across different spectrum domains. The main issue is closing the semantic gap among faces ...

read more

Selected Publications

“Deep learning for understanding faces: machines may be just as good, or better, than humans”

R. Ranjan, S. Sankaranarayanan, A. Bansal, N. Bodla, J-C. Chen, V. M. Patel, C. D. Castillo and R. Chellappa

“HyperFace: a deep multi-task learning framework for face detection, landmark localization, pose estimation, and gender recognition”

R. Ranjan, V. M. Patel, and R. Chellappa

“Unconstrained still/video-based face verification with deep convolutional neural networks”

J-C. Chen, R. Ranjan, S. Sankaranarayanan, A. Kumar, C-H. Chen, V. M. Patel, C. D. Castillo, and R. Chellappa

“The robust sparse Fourier transform and its application in radar signal processing”

S. Wang, V. M. Patel and A. Petropulu
View all Publications