Results and Analysis of the ChaLearn Gesture Challenge 2012

Guyon, Isabelle; Athitsos, Vassilis; Jangyodsuk, Pat; Hamner, Ben

doi:10.1007/978-3-642-40303-3_19

Cited by 44 publications

(39 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the case of TV broadcasts the weak supervision is provided by aligned subtitles (that specify a temporal interval where the word may occur), though the supervision is also noisy as the subtitle word may not be signed. In the case of linguistic research datasets (and some gesture datasets [19]) the supervision is often at the video clip level, rather than a tighter temporal interval.…”

Section: Discussionmentioning

confidence: 99%

“…This is inherently expensive and does not scale to large, evolving gesture languages with high levels of variation. As a result, several recent works have attempted to learn gestures at the other extreme -from single training examples using one-shot learning [16,17,19,20,22,24,33]. However, given the vast variability in how gestures are performed, and the variation in people and camera viewpoints, learning accurate, generalizable models with so little supervision is somewhat challenging, to say the least.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Domain-Adaptive Discriminative One-Shot Learning of Gestures

Pfister

Charles

Zisserman

2014

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The objective of this paper is to recognize gestures in videos -both localizing the gesture and classifying it into one of multiple classes. We show that the performance of a gesture classifier learnt from a single (strongly supervised) training example can be boosted significantly using a 'reservoir' of weakly supervised gesture examples (and that the performance exceeds learning from the one-shot example or reservoir alone). The one-shot example and weakly supervised reservoir are from different 'domains' (different people, different videos, continuous or noncontinuous gesturing, etc.), and we propose a domain adaptation method for human pose and hand shape that enables gesture learning methods to generalise between them. We also show the benefits of using the recently introduced Global Alignment Kernel [12], instead of the standard Dynamic Time Warping that is generally used for time alignment. The domain adaptation and learning methods are evaluated on two large scale challenging gesture datasets: one for sign language, and the other for Italian hand gestures. In both cases performance exceeds the previous published results, including the best skeleton-classification-only entry in the 2013 ChaLearn challenge.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Domain-Adaptive Discriminative One-Shot Learning of Gestures

Pfister

Charles

Zisserman

2014

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…HMMs and CRFs are widely applied for sequential gesture recognition [7]. Another approach is to start from static posture recognition and to apply standard methods used for unsequential data, such as SVMs, on the frame level, and subsequently combine the results into the sequence level.…”

Section: Related Workmentioning

confidence: 99%

Online RGB-D gesture recognition with extreme learning machines

Chen

Koskela

2013

Proceedings of the 15th ACM on International Conference on Multimodal Interaction

View full text Add to dashboard Cite

Gesture recognition is needed in many applications such as human-computer interaction and sign language recognition. The challenges of building an actual recognition system do not lie only in reaching an acceptable recognition accuracy but also with requirements for fast online processing. In this paper, we propose a method for online gesture recognition using RGB-D data from a Kinect sensor. Frame-level features are extracted from RGB frames and the skeletal model obtained from the depth data, and then classified by multiple extreme learning machines. The outputs from the classifiers are aggregated to provide the final classification results for the gestures. We test our method on the ChaLearn multi-modal gesture challenge data. The results of the experiments demonstrate that the method can perform effective multi-class gesture recognition in real-time.

show abstract

“…Kinect was used because of its ability to gather color and depth information from video streams. For 2 years, a vast data set (CGD11), of both development and validation batches, was used worldwide as training and testing data in the competition; the results for both years were reported by Guyon et al (2012Guyon et al ( , 2013 with partial success being achieved. Among the results reported, the Levenshtein distance (LD) (Levenshtein, 1966) was between 0.15 and 0.3 (the ideal distance is represented by 0 and the worst by 1).…”

Section: One-shot Learningmentioning

confidence: 99%

A Human-Centered Approach to One-Shot Gesture Learning

Cabrera

Wachs

2017

Front. Robot. AI

View full text Add to dashboard Cite

This article discusses the problem of one-shot gesture recognition using a humancentered approach and its potential application to fields such as human-robot interaction where the user's intentions are indicated through spontaneous gesturing (one shot). Casual users have limited time to learn the gestures interface, which makes one-shot recognition an attractive alternative to interface customization. In the aim of natural interaction with machines, a framework must be developed to include the ability of humans to understand gestures from a single observation. Previous approaches to oneshot gesture recognition have relied heavily on statistical and data-mining-based solutions and have ignored the mechanisms that are used by humans to perceive and execute gestures and that can provide valuable context information. This omission has led to suboptimal solutions. The focus of this study is on the process that leads to the realization of a gesture, rather than on the gesture itself. In this case, context involves the way in which humans produce gestures-the kinematic and anthropometric characteristics. In the method presented here, the strategy is to generate a data set of realistic samples based on features extracted from a single gesture sample. These features, called the "gist of a gesture," are considered to represent what humans remember when seeing a gesture and, later, the cognitive process involved when trying to replicate it. By adding meaningful variability to these features, a large training data set is created while preserving the fundamental structure of the original gesture. The availability of a large data set of realistic samples allows the use of training classifiers for future recognition. The performance of the method is evaluated using different lexicons, and its efficiency is compared with that of traditional N-shot learning approaches. The strength of the approach is further illustrated through human and machine recognition of gestures performed by a dual-arm robotic platform.

show abstract

Results and Analysis of the ChaLearn Gesture Challenge 2012

Cited by 44 publications

References 14 publications

Domain-Adaptive Discriminative One-Shot Learning of Gestures

Domain-Adaptive Discriminative One-Shot Learning of Gestures

Online RGB-D gesture recognition with extreme learning machines

A Human-Centered Approach to One-Shot Gesture Learning

Contact Info

Product

Resources

About