Mid-Atlantic Computer Vision (MACV) Workshop 2016

May 6, 2016, 8:45AM - 6:00pm

Hodson Hall 110, Johns Hopkins University

Overview

Welcome to the third meeting of the Mid-Atlantic Computer Vision workshop. The 2016 edition of this workshop is organized by Rene Vidal, Alan Yuille, and Austin Reiter. We bring together computer vision researchers around the Mid-Atlantic region of the United States for an informal exchange of ideas in all areas of computer vision. We welcome people to present new research, work that will be presented at upcoming conferences, reviews of not-well-publicized work, and descriptions of "work in progress." We hope that these discussions will let everyone get feedback on their work, inspire new research directions, and maybe even kindle new collaborations.



Location

The workshop will be held in Hodson Hall 110 at Johns Hopkins University. It is located at 3101 Wyman Park Dr, Baltimore, MD 21218.



Registration

Registration for MACV 2016 is free and you can see the list (and register) here.



Hotel and Parking Information

For hotel reservations, we recommend the Inn at The Colonnade Baltimore, which is walking distance from the workshop venue.


Visitors to Johns Hopkins University should park in the South Garage, San Martin Garage, the North Visitor Lot, or metered areas.  Parking in any other area, lot, roadway, or campus property may result in ticketing, booting, and/or towing.  More info on the JHU parking options can be found here.



Talks

Please keep presentations to 12 minutes talk + 3 min questions in length.



Posters

Please create your posters in either ARCH D (24” x 35”) or ARCH E (35” x 48”).



Schedule

Time                                  

8:45-9:00            Opening Session


Oral Session: Object Detection and PCA

9:00-9:15            Human Parsing and Symbiotic Object Detection and Parsing - Jun Zhu, Johns Hopkins

9:15-9:30            SSD: Single Shot MultiBox Detector- Wei Liu, UNC

9:30-9:45            G-CNN: an Iterative Grid Based Object Detector - Mahyar Najibi, UMD

9:45-10:00          Closing the Barn Door: Fast Detection of Intruders on Mobile Devices - Pramuditha Perera, Rutgers

10:00-10:15        Learning Scene-Specific Pedestrian Detectors without Real Data - Vishnu Boddeti, CMU

10:15-10:30        Nonparametric Additive Component Analysis - Calvin Murdock, CMU


10:30-11:00        Coffee Break


Oral Session: Segmentation and Clustering

11:00-11:15        Scalable Elastic Net Subspace Clustering - Chong You, JHU

11:15-11:30        Gaussian CRF Network for Semantic Segmentation - Raviteja Vemulapalli, UMD

11:30-11:45        Joint Semantic Segmentation and Depth Estimation with Deep Convolutional Networks - Arsalan            

                           Mousavian, George Mason


Oral Session: Visual Questions and Language

11:45-12:00        On Efficient Bayesian Scene Interpretation: An Entropy Pursuit Approach - Ehsan Jahangiri, JHU

12:00-12:15        Yin and Yang: Balancing and Answering Binary Visual Questions - Yash Goyal, Virginia Tech

12:15-12:30        Teaching Machines to See with Language - Mohamed ELhoseiny, Rutgers


12:30-2:00           Lunch Break


Oral Session: 3D Computer Vision

2:00-2:15            Marr Revisited: 2D-3D Alignment via Surface Normal Prediction - Aayush Bansal, CMU

2:15-2:30            On the Fall and Rise of Direct Methods - Hatem Alismail, CMU

2:30-2:45            Multi-domain pooling and its application in 3D perception - Chi Li, JHU

2:45-3:00            3D Deep Learning for Robot Perception - Fisher Yu, Princeton

3:00-3:15            Active Vision for Cognitive Robots - Yezhou Yang, UMD

3:15-3:30            Monocular 3D Object Reconstruction - Xiaowei Zhou, UPenn


3:30-4:30            Poster Session and Coffee Break


Oral Session: Actions, Humor, Privacy and Tracking

4:30-4:45            Improving Human Action Recognition by Non-action Classification - Minh Hoai Nguyen, Stony Brook

4:45-5:00            Segmental Spatio-Temporal CNNs for Fine-grained Action Segmentation - Colin Lea, JHU

5:00-5:15            Learning Action Maps of Large Environments via First Person Vision - Nick Rhinehart, CMU

5:15-5:30            We Are Humor Beings: Understanding and Predicting Visual Humor - Arjun Chandrasekaran, Virginia

                           Tech

5:30-5:45            Computer Vision for Data Privacy - Yi Xu, UNC

5:45-6:00            Performance Evaluation of Multi-Target, Multi-Camera Tracking Systems - Ergys Ristani, Duke



Wireless Internet on Campus

You can connect to Hopkins-Guest from your laptop easily.



Contact Information

If you have any further questions or comments about the website, please contact, Rene Vidal, at rvidal-at-cis.jhu.edu (replacing -at- with @).






































© 2016 Johns Hopkins University