H
Heesung Kwon
Researcher at United States Army Research Laboratory
Publications - 146
Citations - 3776
Heesung Kwon is an academic researcher from United States Army Research Laboratory. The author has contributed to research in topics: Object detection & Kernel embedding of distributions. The author has an hindex of 23, co-authored 134 publications receiving 2746 citations. Previous affiliations of Heesung Kwon include University at Buffalo & University of Massachusetts Amherst.
Papers
More filters
Posted Content
DOD-CNN: Doubly-injecting Object Information for Event Recognition
TL;DR: Doubly-injected object detection CNN (DOD-CNN) as mentioned in this paper exploits the object information in both ways for the task of event recognition by detecting relevant objects in two ways: (i) indirectly utilizing object detection information within the unified architecture or (ii) directly making use of the object detection output results.
Posted Content
Discovering and Generating Hard Examples for Training a Red Tide Detector.
TL;DR: The proposed 9-layer fully convolutional network jointly trained with cOHEM and HNG provides state-of-the-art red tide detection accuracy on GOCI satellite images.
Proceedings ArticleDOI
Negative Samples are at Large: Leveraging Hard-distance Elastic Loss for Re-identification
TL;DR: MoReID as discussed by the authors uses a dictionary to store current and past batches to build a large set of encoded samples to leverage a very large number of negative samples in training for general re-ID task.
Posted Content
Dynamic Belief Fusion for Object Detection
TL;DR: In this article, a novel approach for the fusion of heterogeneous object detection methods is proposed, called Dynamic Belief Fusion (DBF), which dynamically assigns probabilities to hypotheses based on confidence levels in the detection results conditioned on the prior performance of individual detectors.
Journal ArticleDOI
A novel convolutional neural network for deep-learning classification
Jared Shamwell,Hyungtae Lee,Heesung Kwon,Amar R. Marathe,Vernon J. Lawhern,William D. Nothwang +5 more
TL;DR: This work has shown that machine learning approaches can be applied to raw human neurophysiological data and thus reveal signals that are relevant to the detection of target images and can increase both the accuracy and the response rate of image triage classification tasks.