scispace - formally typeset
Search or ask a question
Author

Eric Lecolinet

Other affiliations: Université Paris-Saclay, École Normale Supérieure, ParisTech  ...read more
Bio: Eric Lecolinet is an academic researcher from Télécom ParisTech. The author has contributed to research in topics: Gesture & Interaction technique. The author has an hindex of 29, co-authored 139 publications receiving 3565 citations. Previous affiliations of Eric Lecolinet include Université Paris-Saclay & École Normale Supérieure.


Papers
More filters
Journal ArticleDOI
TL;DR: H holistic approaches that avoid segmentation by recognizing entire character strings as units are described, including methods that partition the input image into subimages, which are then classified.
Abstract: Character segmentation has long been a critical area of the OCR process. The higher recognition rates for isolated characters vs. those obtained for words and connected character strings well illustrate this fact. A good part of recent progress in reading unconstrained printed and written text may be ascribed to more insightful handling of segmentation. This paper provides a review of these advances. The aim is to provide an appreciation for the range of techniques that have been developed, rather than to simply list sources. Segmentation methods are listed under four main headings. What may be termed the "classical" approach consists of methods that partition the input image into subimages, which are then classified. The operation of attempting to decompose the image into classifiable units is called "dissection." The second class of methods avoids dissection, and segments the image either explicitly, by classification of prespecified windows, or implicitly by classification of subsets of spatial features collected from the image as a whole. The third strategy is a hybrid of the first two, employing dissection together with recombination rules to define potential segments, but using classification to select from the range of admissible segmentation possibilities offered by these subimages. Finally, holistic approaches that avoid segmentation by recognizing entire character strings as units are described.

880 citations

Proceedings ArticleDOI
28 May 2008
TL;DR: The design and evaluation of TapTap and MagStick are presented, two thumb interaction techniques for target acquisition on mobile devices with small touch-screens that are found to be faster than four previous techniques except Direct Touch which, although faster, is too error prone.
Abstract: We present the design and evaluation of TapTap and MagStick, two thumb interaction techniques for target acquisition on mobile devices with small touch-screens. These two techniques address all the issues raised by the selection of targets with the thumb on small tactile screens: screen accessibility, visual occlusion and accuracy. A controlled experiment shows that TapTap and MagStick allow the selection of targets in all areas of the screen in a fast and accurate way. They were found to be faster than four previous techniques except Direct Touch which, although faster, is too error prone. They also provided the best error rate of all tested techniques. Finally the paper also provides a comprehensive study of various techniques for thumb based touch-screen target selection.

180 citations

Proceedings ArticleDOI
10 Apr 2010
TL;DR: Experimental evaluations suggest that these two novel navigation techniques, CycloPan and CycloZoom+, for clutch-free 2D panning and browsing, outperform flicking and rubbing techniques.
Abstract: This paper introduces two novel navigation techniques, CycloPan, for clutch-free 2D panning and browsing, and CycloZoom+, for integrated 2D panning and zooming. These techniques instantiate a more generic concept which we call Cyclo* (CycloStar). The basic idea is that users can exert closed-loop control over several continuous variables by voluntarily modulating the parameters of a sustained oscillation. Touch-sensitive surfaces tend to offer impoverished input resources. Cyclo* techniques seem particularly promising on these surfaces because oscillations have multiple geometrical and kinematic parameters many of which may be used as controls. While CycloPan and CycloZoom+ are compatible with each other and with much of the state of the art, our experimental evaluations suggest that these two novel techniques outperform flicking and rubbing techniques.

149 citations

Proceedings ArticleDOI
04 Apr 2009
TL;DR: This work proposes to discriminate, among thumb gestures, those it calls MicroRolls, characterized by zero tangential velocity of the skin relative to the screen surface, and shows that at least 16 elemental gestures can be automatically recognized.
Abstract: The input vocabulary for touch-screen interaction on handhelds is dramatically limited, especially when the thumb must be used. To enrich that vocabulary we propose to discriminate, among thumb gestures, those we call MicroRolls, characterized by zero tangential velocity of the skin relative to the screen surface. Combining four categories of thumb gestures, Drags, Swipes, Rubbings and MicroRolls, with other classification dimensions, we show that at least 16 elemental gestures can be automatically recognized. We also report the results of two experiments showing that the roll vs. slide distinction facilitates thumb input in a realistic copy and paste task, relative to existing interaction techniques.

143 citations

Proceedings ArticleDOI
27 Apr 2013
TL;DR: WatchIt is a prototype device that extends interaction beyond the watch surface to the wristband, and two interaction techniques for command selection and execution, and a novel gesture technique and an adaptation of an existing menu technique suitable for wristband interaction are proposed.
Abstract: We present WatchIt, a prototype device that extends interaction beyond the watch surface to the wristband, and two interaction techniques for command selection and execution. Because the small screen of wristwatch computers suffers from visual occlusion and the fat finger problem, we investigated the use of the wristband as an available interaction resource. Not only does WatchIt use a cheap, energy efficient and invisible technology, but it involves simple, basic gestures that allow good performance after little training, as suggested by the results of a pilot study. We propose a novel gesture technique and an adaptation of an existing menu technique suitable for wristband interaction. In a user study, we investigated their usage in eyes-free contexts, finding that they perform well. Finally, we present techniques where the bracelet is used in addition to the screen to provide precise continuous control over list scrolling. We also report on a preliminary survey of traditional and digital jewelry that points to the high frequency of watches and bracelets in both genders and gives a sense of the tasks people feel like performing on such devices.

138 citations


Cited by
More filters
Journal ArticleDOI
TL;DR: The nature of handwritten language, how it is transduced into electronic data, and the basic concepts behind written language recognition algorithms are described.
Abstract: Handwriting has continued to persist as a means of communication and recording information in day-to-day life even with the introduction of new technologies. Given its ubiquity in human transactions, machine recognition of handwriting has practical significance, as in reading handwritten notes in a PDA, in postal addresses on envelopes, in amounts in bank checks, in handwritten fields in forms, etc. This overview describes the nature of handwritten language, how it is transduced into electronic data, and the basic concepts behind written language recognition algorithms. Both the online case (which pertains to the availability of trajectory data during writing) and the off-line case (which pertains to scanned images) are considered. Algorithms for preprocessing, character and word recognition, and performance with practical systems are indicated. Other fields of application, like signature verification, writer authentification, handwriting learning tools are also considered.

2,653 citations

Journal ArticleDOI
TL;DR: H holistic approaches that avoid segmentation by recognizing entire character strings as units are described, including methods that partition the input image into subimages, which are then classified.
Abstract: Character segmentation has long been a critical area of the OCR process. The higher recognition rates for isolated characters vs. those obtained for words and connected character strings well illustrate this fact. A good part of recent progress in reading unconstrained printed and written text may be ascribed to more insightful handling of segmentation. This paper provides a review of these advances. The aim is to provide an appreciation for the range of techniques that have been developed, rather than to simply list sources. Segmentation methods are listed under four main headings. What may be termed the "classical" approach consists of methods that partition the input image into subimages, which are then classified. The operation of attempting to decompose the image into classifiable units is called "dissection." The second class of methods avoids dissection, and segments the image either explicitly, by classification of prespecified windows, or implicitly by classification of subsets of spatial features collected from the image as a whole. The third strategy is a hybrid of the first two, employing dissection together with recombination rules to define potential segments, but using classification to select from the range of admissible segmentation possibilities offered by these subimages. Finally, holistic approaches that avoid segmentation by recognizing entire character strings as units are described.

880 citations

Journal ArticleDOI
TL;DR: A state-of-the-art review of the development of spiking neurons and SNNs is presented, and insight into their evolution as the third generation neural networks is provided.
Abstract: Most current Artificial Neural Network (ANN) models are based on highly simplified brain dynamics. They have been used as powerful computational tools to solve complex pattern recognition, function estimation, and classification problems. ANNs have been evolving towards more powerful and more biologically realistic models. In the past decade, Spiking Neural Networks (SNNs) have been developed which comprise of spiking neurons. Information transfer in these neurons mimics the information transfer in biological neurons, i.e., via the precise timing of spikes or a sequence of spikes. To facilitate learning in such networks, new learning algorithms based on varying degrees of biological plausibility have also been developed recently. Addition of the temporal dimension for information encoding in SNNs yields new insight into the dynamics of the human brain and could result in compact representations of large neural networks. As such, SNNs have great potential for solving complicated time-dependent pattern recognition problems because of their inherent dynamic representation. This article presents a state-of-the-art review of the development of spiking neurons and SNNs, and provides insight into their evolution as the third generation neural networks.

694 citations