A research scientist in machine learning and computer vision, I strive to understand the mathematical structure of the research questions in order to develop effective and efficient algorithmic solutions, with strong analytical properties and compelling practical performance.

Slides from recent talks

Towards Visual Recognition in the Wild: Long-Tailed Sources and Open Compound Targets [Slides, 30 mins] [Slides, 45 mins]

Presented at CVPR 2020 Workshops on Learning from Imperfect Data and Adversarial Machine Learning in Computer Vision, respectively

Long-Tailed Visual Recognition is A Domain Adaptation Problem [Slides] [PDF]

Presented at CVPR 2020 Area Chair Workshop and WACV 2020 Workshop on Vision Applications and Solutions to Biased or Scarce Data

Gaussian Attack by Learning the Distributions of Adversarial Examples [Slides]

Presented at UC Davis, Waymo, and CVPR 2019 Workshop on GigaVision

Sequential Determinantal Point Processes: Models, Algorithms, and Applications in Diverse and Sequential Subset Selection [Slides]

Curriculum Domain Adaptation: Using Simulation for Real [PDF]

Presented at IEEE BIGDATA 2018 Workshop on Big Data Transfer Learning

Multiple Shades of DropoutĀ  for Discriminative and Generative Deep Neural Networks [PDF]

Presented at INFORMS 2018 special session on Stochastic Optimization Methods and Approximation Theory in Machine Learning

Learning and Adapting from the Web for Visual Recognitiona [PDF]

Presented at CVPR 2018 Workshop on Web Vision and ECCV 2018 Workshop on CEFRL

Domain Adaptation and Transfer: All You Need to Use Simulation "for Real"[PDF]

Presented at ECCV 2018 Workshop on Visual Learning and Embodied Agents in Simulation

Sequential Determinantal Point Processes (SeqDPPs) and Variations for Supervised Video Summarization[PDF]

Presented at UC Berkeley, UC Irvine, Stanford University, Adobe Research, Facebook, etc. in 2017

Domain Adaptation for Robust Visual Recognition [PDF]

Presented at ICSI/UC Berkeley, UC Santa Cruz, Google Research, NVIDIA Research, NEC Labs, etc. in 2017

Recent Reseach

Some of the topics I have been working on over the recent couple of years.

Domain Adaptation

with applications to object recognition,
face detection, semantic segmentation,
acition recognition, and attribute detection

Sequential Determinantal Point Processes

with applications to video summarization
and multi-documment summarization

Vision & Language

such as visual question answering,
video captioning, and image tagging

Data Efficient Learning

such as zero-/low-shot learning, semi-
supervised learning, Webly-supervised
learning, and domain adaptation

Reinforcement Learning

with applications to video summarization,
visual navigation, etc.

3D objects & faces

3D object retrieval, 3D face recognition,
and attribute detection of 3D objects

Reseach Grants

My research was supported by the following grants.

NSF BIGDATA: IA: Distributed Semi-Supervised Training of Deep Models and Its Applications in Video Understanding

Funding bodies: National Science Foundation
Role: Principle investigator
Value: 1/3 of ($662,431 + $42,500 AWS Credits)
Years: 2017-2020

NSF CRII: RI: Multi-Source Domain Generalization Approaches to Visual Attribute Detection

Funding bodies: National Science Foundation
Role: Sole Principle investigator
Value: $175,000
Years: 2016-2018

Face Detector Adaptation without Forgetting

Funding bodies: Adobe Systems Inc
Role: Sole Principle investigator
Value: $10,000
Years: 2017

Multiple-Modal Summarization of Videos and Photo Albums with User Input

Funding bodies: FutureWei Technologies Inc
Role: Sole Principle investigator
Value: $100,000
Years: 2017

User-Guided Visual Analytics

Funding bodies: Adobe Systems Inc
Role: Sole Principle investigator
Value: $7,000
Years: 2016


List of Publications

Work hard, make extraordinary discoveries, and publish to make the work simple.