• Publications
  • Influence
Joint Detection and Recounting of Abnormal Events by Learning Deep Generic Knowledge
TLDR
This paper addresses the problem of joint detection and recounting of abnormal events in videos by integrating a generic CNN model and environment-dependent anomaly detectors. Expand
  • 80
  • 13
  • PDF
Efficient Image Retrieval via Decoupling Diffusion into Online and Offline Processing
TLDR
We propose to use late truncation instead of early truncation in previous works to improve the efficiency of the diffusion process to achieve better retrieval performance. Expand
  • 14
  • 5
  • PDF
Region-Based Image Retrieval Revisited
TLDR
We propose a multitask CNN feature that allows us to use deep learning technique and to jointly handle multi-aspect object specification. Expand
  • 12
  • 1
  • PDF
Discriminative Learning of Open-Vocabulary Object Retrieval and Localization by Negative Phrase Augmentation
TLDR
We propose Query-Adaptive R-CNN, a simple extension of Faster R- CNN adapted to open-vocabulary queries, by transforming the text embedding vector into an object classifier and localization regressor. Expand
  • 11
  • 1
  • PDF
Query-Adaptive R-CNN for Open-Vocabulary Object Detection and Retrieval
TLDR
We address the problem of open-vocabulary object retrieval and localization, which is to retrieve and localize objects from a large-scale image database immediately by a textual query (e.g., a word or phrase). Expand
  • 2
  • 1
Large-Scale R-CNN with Classifier Adaptive Quantization
TLDR
This paper extends R-CNN, a state-of-the-art object detection method, to larger scales. Expand
  • 6
  • PDF
Audience Behavior Mining: Integrating TV Ratings with Multimedia Content
TLDR
We propose a framework for mining audience behavior from TV ratings data by combining the filters interactively. Expand
  • 5
  • PDF
Audience Behavior Mining by Integrating TV Ratings with Multimedia Contents
TLDR
This paper presents a framework that enables discovery of numerous patterns of audience behavior from TV ratings. Expand
  • 3
  • PDF
Multimodal Co-Training for Selecting Good Examples from Webly Labeled Video
TLDR
We tackle the problem of learning concept classifiers from videos on the web without using manually labeled data. Expand
  • 4
  • PDF
Reconfigurable Inverted Index
TLDR
We develop a reconfigurable inverted index (Rii) to resolve these two fundamental problems that have not received sufficient attention from the research community. Expand
  • 1
  • PDF