arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Thomas Stepleton

11 papers · 2002–2018

2002: 12003: 02004: 02005: 12006: 12007: 12008: 22009: 02010: 12011: 02012: 02013: 22014: 02015: 02016: 02017: 02018: 2
2002peak 2 in 20082018

Frequent co-authors

Papers

showing 1–11 of 11 RSS

Iterative design of a Braille writing tutor to combat illiteracy

Nidhi Kalra, Tom Lauwers, D. Dewey +2 more

Human Factors and Ergonomics · Figshare · · cited 46× · DOI ↗ · more like this ·

Less than 3% of the 145 million blind people living in developing countries are literate. This low literacy rate is partly due to the lack of trained teachers and the challenges associated with learning to write Braille on a traditional slate and stylus. These challenges include writing from right to left, writing mirrored images of letters, and receiving significantly delayed feedback. Extensive conversations with the Mathru School for the Blind near Bangalore, India, revealed the need for a robust, low-power, low-cost Braille writing tutor. We present an iterative and participatory design process resulting in the creation and refinement of a prototype Braille writing tutor system. This system uses a novel input device to capture a student's activity on a slate using a stylus and uses a range of techniques to teach Braille writing skills to both beginner and advanced students. We…

Towards unsupervised whole-object segmentation: Combining automated matting with boundary detection

Andrew N. Stein, Thomas Stepleton, Martial Hebert

Computer Vision and Pattern Recognition · Research Showcase @ Carnegie Mellon University (Carnegie Mellon University) · · cited 33× · DOI ↗ · more like this ·

We propose a novel step toward the unsupervised segmentation of whole objects by combining ldquohintsrdquo of partial scene segmentation offered by multiple soft, binary mattes. These mattes are implied by a set of hypothesized object boundary fragments in the scene. Rather than trying to find or define a single ldquobestrdquo segmentation, we generate multiple segmentations of an image. This reflects contemporary methods for unsupervised object discovery from groups of images, and it allows us to define intuitive evaluation metrics for our sets of segmentations based on the accurate and parsimonious delineation of scene objects. Our proposed approach builds on recent advances in spectral clustering, image matting, and boundary detection. It is demonstrated qualitatively and quantitatively on a dataset of scenes and is suitable for current work in unsupervised object discovery without…

The Block Diagonal Infinite Hidden Markov Model

Thomas Stepleton, Zoubin Ghahramani, Geoffrey J. Gordon +1 more

Signal Processing · Figshare · · cited 15× · DOI ↗ · more like this ·

The Infinite Hidden Markov Model (IHMM) extends hidden Markov models to have a countably infinite number of hidden states . We present a generalization of this framework that introduces block-diagonal structure in the transitions between the hidden states. These blocks correspond to "sub-behaviors" exhibited by data sequences. In identifying such structure, the model classifies, or partitions, sequence data according to these sub-behaviors in an unsupervised way. We present an application of this model to artificial data, a video gesture classification task, and a musical theme labeling task, and show that components of the model can also be applied to graph segmentation.

Work-Augmented Laziness with the Los Task Request System

Thomas Stepleton

Information Systems · USENIX Large Installation Systems Administration Conference · · cited 3× · DOI ↗ · more like this ·

Endothelial-to-mesenchymal transition (EndMT) mainly exists in cardiovascular development and disease progression, and is well known to contribute to cardiac fibrosis. Recent studies indicated that autophagy also participates in the regulation of cardiac fibrosis. However, the precise role of autophagy in cardiac fibrosis and the underlying molecular mechanism remain unclear. The present study aimed to explore the role of autophagy in EndMT, reveal the underlying molecular mechanism, and seek new therapy for cardiac fibrosis. In the present study, we found that EndMT and autophagy were induced simultaneously by hypoxia in human cardiac microvascular endothelial cells (HCMECs). Rapamycin, an autophagy enhancer, attenuated EndMT with promoting angiogenesis, while 3-methyladenine (3-MA) and chloroquine (CQ), agents that inhibit autophagy, accelerated the progression accompanied by the…

Low-pass Recurrent Neural Networks - A memory architecture for longer-term correlation discovery

Thomas Stepleton, Razvan Pascanu, Will Dabney +3 more

cs.LG · · cited 2× · arXiv ↗ · more like this ·

Reinforcement learning (RL) agents performing complex tasks must be able to remember observations and actions across sizable time intervals. This is especially true during the initial learning stages, when exploratory behaviour can increase the delay between specific actions and their effects. Many new or popular approaches for learning these distant correlations employ backpropagation through time (BPTT), but this technique requires storing observation traces long enough to span the interval between cause and effect. Besides memory demands, learning dynamics like vanishing gradients and slow convergence due to infrequent weight updates can reduce BPTT's practicality; meanwhile, although online recurrent network learning is a developing topic, most approaches are not efficient enough to use as replacements. We propose a simple, effective memory strategy that can extend the window over…

Toward versatile structural modification for bayesian nonparametric time series models

Tai Sing Lee, Thomas Stepleton

Artificial Intelligence · · cited 1× · OpenAlex ↗ · more like this ·

Unsupervised learning techniques discover organizational structure in data, but to do so they must approach the problem with a priori assumptions. A fundamental trend in the development of these techniques has been the relaxation or elimination of the unwanted or arbitrary structural assumptions they impose. For systems that derive hidden Markov models (HMMs) from time series data, state-of-the-art techniques now assume only that the number of hidden states will be relatively small, a useful, flexible, and usually correct hypothesis. With unwanted structural constraints mitigated, we investigate a flexible means of introducing new, useful structural assumptions into an advanced HMM learning technique, assumptions that reflect details of our prior understanding of the problem. Our investigation, motivated by the unsupervised learning of view-based object models from video data, adapts…

Using Co-Occurrence and Segmentation to Learn Feature-Based Object Models from Video

Thomas Stepleton, Tai Sing Lee

Computer Vision and Pattern Recognition · Research Showcase @ Carnegie Mellon University (Carnegie Mellon University) · · cited 1× · DOI ↗ · more like this ·

A number of recent systems for unsupervised feature- based learning of object models take advantage of cooccurrence: broadly, they search for clusters of discriminative features that tend to coincide across multiple still images or video frames. An intuition behind these efforts is that regularly co-occurring image features are likely to refer to physical traits of the same object, while features that do not often co-occur are more likely to belong to different objects. In this paper we discuss a refinement to these techniques in which multiple segmentations establish meaningful contexts for co-occurrence, or limit the spatial regions in which two features are deemed to co-occur. This approach can reduce the variety of image data necessary for model learning and simplify the incorporation of less discriminative features into the model.

Recommended Citation

Thomas Stepleton, Zoubin Ghahramani, Geoffrey J. Gordon +5 more

Computer Vision and Pattern Recognition · · cited 0× · OpenAlex ↗ · more like this ·

The Infinite Hidden Markov Model (IHMM) extends hidden Markov models to have a countably infinite number of hidden states (Beal et al., 2002; Teh et al., 2006). We present a generalization of this framework that introduces nearly block-diagonal structure in the transitions between the hidden states, where blocks correspond to “subbehaviors” exhibited by data sequences. In identifying such structure, the model classifies, or partitions, sequence data according to these sub-behaviors in an unsupervised way. We present an application of this model to artificial data, a video gesture classification task, and a musical theme labeling task, and show that components of the model can also be applied to graph segmentation. 1

Combining Automated Matting with Boundary Detection

Andrew N. Stein, Thomas S. Stepleton

Computer Vision and Pattern Recognition · · cited 0× · OpenAlex ↗ · more like this ·

We propose a novel step toward the unsupervised segmentation of whole objects by combining “hints ” of partial scene segmentation offered by multiple soft, binary mattes. These mattes are implied by a set of hypothesized object boundary fragments in the scene. Rather than trying to find or define a single “best ” segmentation, we generate multiple segmentations of an image. This reflects contemporary methods for unsupervised object discovery from groups of images, and it allows us to define intuitive evaluation metrics for our sets of segmentations based on the accurate and parsimonious delineation of scene objects. Our proposed approach builds on recent advances in spectral clustering, image matting, and boundary detection. It is demonstrated qualitatively and quantitatively on a dataset of scenes and is suitable for current work in unsupervised object discovery without top-down…

Toward Unsupervised Whole-Object Segmentation: Combining Automated Matting with Boundary Detection

Andrew N. Stein, Thomas S. Stepleton

Artificial Intelligence · · cited 0× · OpenAlex ↗ · more like this ·

We propose a novel step toward the unsupervised seg-mentation of whole objects by combining “hints ” of partial scene segmentation offered by multiple soft, binary mattes. These mattes are implied by a set of hypothesized object boundary fragments in the scene. Rather than trying to find or define a single “best ” segmentation, we generate multi-ple segmentations of an image. This reflects contemporary methods for unsupervised object discovery from groups of images, and it allows us to define intuitive evaluation met-rics for our sets of segmentations based on the accurate and parsimonious delineation of scene objects. Our proposed approach builds on recent advances in spectral clustering, image matting, and boundary detection. It is demonstrated qualitatively and quantitatively on a dataset of scenes and is suitable for current work in unsupervised object discovery without top-down…

Predicting and Evaluating the Power of Shared Features

Thomas Stepleton

Computer Vision and Pattern Recognition · · cited 0× · DOI ↗ · more like this ·

Several recent efforts in multi-class feature-based object recognition employ shared features, or features that simultaneously belong to multiple class models. These approaches claim a considerable time savings by reducing the total number of features used by all models, thereby lessening the concomitant computational effort of finding the features in images. In this paper we derive a Bayesian framework for predicting and evaluating the performance of shared feature-based recognition systems. We then use this framework to predict the performance of several instances of a simple multi-class object detector.