Kevin Lin

Kevin Lin

Auto-generated profile, based on publicly available information

Kevin Lin

University of Arizona

Research Focus & Metrics

Academic Career Length

19Years (2008-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 June 2015
Verified
Conference Paper
Closed access
KLKevin Lin
Kevin Lin
HYHuei-Fang Yang
Huei-Fang Yang
JHJen-Hao Hsiao
Jen-Hao Hsiao
CCChu-Song Chen
Chu-Song Chen
Approximate nearest neighbor search is an efficient strategy for large-scale image retrieval. Encouraged by the recent advances in convolutional neural networks (CNNs), we propose an effective deep learning framework to generate binary hash codes for fast image retrieval. Our idea is that when the data labels are available, binary codes can be learned by employing a hidden layer for representing the latent concepts that dominate the class labels. The utilization of the CNN also allows for learning image representations. Unlike other supervised methods that require pair-wised inputs for binary code learning, our method learns hash codes and image representations in a point-wised manner, making it suitable for large-scale datasets. Experimental results show that our method outperforms several state-of-the-art hashing algorithms on the CIFAR-10 and MNIST datasets. We further demonstrate its scalability and efficacy on a large-scale dataset of 1 million clothing images.

1 June 2016
Verified
Conference Paper
Closed access
KLKevin Lin
Kevin Lin
JLJiwen Lu
Jiwen Lu
CCChu-Song Chen
Chu-Song Chen
JZJie Zhou
Jie Zhou
In this paper, we propose a new unsupervised deep learning approach called DeepBit to learn compact binary descriptor for efficient visual object matching. Unlike most existing binary descriptors which were designed with random projections or linear hash functions, we develop a deep neural network to learn binary descriptors in an unsupervised manner. We enforce three criterions on binary codes which are learned at the top layer of our network: 1) minimal loss quantization, 2) evenly distributed codes and 3) uncorrelated bits. Then, we learn the parameters of the networks with a back-propagation technique. Experimental results on three different visual analysis tasks including image matching, image retrieval, and object recognition clearly demonstrate the effectiveness of the proposed approach.

9 February 2017 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Journal Article
Closed access
HYHuei-Fang Yang
Huei-Fang Yang
KLKevin Lin
Kevin Lin
CCChu-Song Chen
Chu-Song Chen
This paper presents a simple yet effective supervised deep hash approach that constructs binary hash codes from labeled data for large-scale image search. We assume that the semantic labels are governed by several latent attributes with each attribute on or off, and classification relies on these attributes. Based on this assumption, our approach, dubbed supervised semantics-preserving deep hashing (SSDH), constructs hash functions as a latent layer in a deep network and the binary codes are learned by minimizing an objective function defined over classification error and other desirable hash codes properties. With this design, SSDH has a nice characteristic that classification and retrieval are unified in a single learning model. Moreover, SSDH performs joint learning of image representations, hash codes, and classification in a point-wised manner, and thus is scalable to large-scale datasets. SSDH is simple and can be realized by a slight enhancement of an existing deep architecture for classification; yet it is effective and outperforms other hashing approaches on several benchmarks and large datasets. Compared with state-of-the-art approaches, SSDH achieves higher retrieval accuracy, while the classification performance is not sacrificed.

7 September 2023 in Nature Biotechnology
Verified
Journal Article
Open access
PDF file
SCShuxiao Chen
Shuxiao Chen
BZBokai Zhu
Bokai Zhu
SHSijia Huang
Sijia Huang
JHJohn W. Hickey
John W. Hickey
KLKevin Z. Lin
Kevin Z. Lin
ZMZongming Ma
Zongming Ma
Although single-cell and spatial sequencing methods enable simultaneous measurement of more than one biological modality, no technology can capture all modalities within the same cell. For current data integration methods, the feasibility of cross-modal integration relies on the existence of highly correlated, a priori ‘linked’ features. We describe matching X-modality via fuzzy smoothed embedding (MaxFuse), a cross-modal data integration method that, through iterative coembedding, data smoothing and cell matching, uses all information in each modality to obtain high-quality integration even when features are weakly linked. MaxFuse is modality-agnostic and demonstrates high robustness and accuracy in the weak linkage scenario, achieving 20~70% relative improvement over existing methods under key evaluation metrics on benchmarking datasets. A prototypical example of weak linkage is the integration of spatial proteomic data with single-cell sequencing data. On two example analyses of this type, MaxFuse enabled the spatial consolidation of proteomic, transcriptomic and epigenomic information at single-cell resolution on the same tissue section.

1 February 2017 in Physica D: Nonlinear Phenomena
Verified
Journal Article
Closed access
FLFei Lu
Fei Lu
KLKevin K. Lin
Kevin K. Lin
ACAlexandre J. Chorin
Alexandre J. Chorin

1 January 2013 in The Journal of Mathematical Neuroscience
Verified
Journal Article
Open access
PDF file
KWKyle Ca Wedgwood
Kyle Ca Wedgwood
KLKevin K Lin
Kevin K Lin
RTRuediger Thul
Ruediger Thul
SCStephen Coombes
Stephen Coombes
Phase oscillators are a common starting point for the reduced description of many single neuron models that exhibit a strongly attracting limit cycle. The framework for analysing such models in response to weak perturbations is now particularly well advanced, and has allowed for the development of a theory of weakly connected neural networks. However, the strong-attraction assumption may well not be the natural one for many neural oscillator models. For example, the popular conductance based Morris–Lecar model is known to respond to periodic pulsatile stimulation in a chaotic fashion that cannot be adequately described with a phase reduction. In this paper, we generalise the phase description that allows one to track the evolution of distance from the cycle as well as phase on cycle. We use a classical technique from the theory of ordinary differential equations that makes use of a moving coordinate system to analyse periodic orbits. The subsequent phase-amplitude description is shown to be very well suited to understanding the response of the oscillator to external stimuli (which are not necessarily weak). We consider a number of examples of neural oscillator models, ranging from planar through to high dimensional models, to illustrate the effectiveness of this approach in providing an improvement over the standard phase-reduction technique. As an explicit application of this phase-amplitude framework, we consider in some detail the response of a generic planar model where the strong-attraction assumption does not hold, and examine the response of the system to periodic pulsatile forcing. In addition, we explore how the presence of dynamical shear can lead to a chaotic response.