PyTorchKR Papers

최신 논문

총 237편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1408.1759

Real-Time and Robust Method for Hand Gesture Recognition System Based on Cross-Correlation Coefficient

Hand gesture recognition possesses extensive applications in virtual reality, sign language recognition, and computer games. The direct interface of hand gestures provides us a new way for communicating with the virtual environment. In this paper a novel and real-time approach for hand gesture recognition system is presented. In the suggested method, first, the hand gesture is extracted from the main image by the image segmentation and morphological operation and then is sent to feature extraction stage. In feature extraction stage the Cross-correlation coefficient is applied on the gesture to recognize it. In the result part, the proposed approach is applied on American Sign Language (ASL) database and the accuracy rate obtained 98.34%.

arXiv:1409.1411

Visual Speech Recognition

Lip reading is used to understand or interpret speech without hearing it, a technique especially mastered by people with hearing difficulties. The ability to lip read enables a person with a hearing impairment to communicate with others and to engage in social activities, which otherwise would be difficult. Recent advances in the fields of computer vision, pattern recognition, and signal processing has led to a growing interest in automating this challenging task of lip reading. Indeed, automating the human ability to lip read, a process referred to as visual speech recognition (VSR) (or sometimes speech reading), could open the door for other novel related applications. VSR has received a great deal of attention in the last decade for its potential use in applications such as human-computer interaction (HCI), audio-visual speech recognition (AVSR), speaker recognition, talking heads, sign language recognition and video surveillance. Its main aim is to recognise spoken word(s) by using only the visual signal that is produced during speech. Hence, VSR deals with the visual domain of speech and involves image processing, artificial intelligence, object detection, pattern recognition, statistical modelling, etc.

arXiv:1503.05830

Sign Language Fingerspelling Classification from Depth and Color Images using a Deep Belief Network

Automatic sign language recognition is an open problem that has received a lot of attention recently, not only because of its usefulness to signers, but also due to the numerous applications a sign classifier can have. In this article, we present a new feature extraction technique for hand pose recognition using depth and intensity images captured from a Microsoft Kinect sensor. We applied our technique to American Sign Language fingerspelling classification using a Deep Belief Network, for which our feature extraction technique is tailored. We evaluated our results on a multi-user data set with two scenarios: one with all known users and one with an unseen user. We achieved 99% recall and precision on the first, and 77% recall and 79% precision on the second. Our method is also capable of real-time sign classification and is adaptive to any environment or lightning intensity.

arXiv:1509.03001공식 코드 · Papers with Code

Real-time Sign Language Fingerspelling Recognition using Convolutional Neural Networks from Depth map

Sign language recognition is important for natural and convenient communication between deaf community and hearing majority. We take the highly efficient initial step of automatic fingerspelling recognition system using convolutional neural networks (CNNs) from depth maps. In this work, we consider relatively larger number of classes compared with the previous literature. We train CNNs for the classification of 31 alphabets and numbers using a subset of collected depth data from multiple subjects. While using different learning configurations, such as hyper-parameter selection with and without validation, we achieve 99.99% accuracy for observed signers and 83.58% to 85.49% accuracy for new signers. The result shows that accuracy improves as we include more data from different subjects during training. The processing time is 3 ms for the prediction of a single image. To the best of our knowledge, the system achieves the highest accuracy and speed. The trained model and dataset is available on our repository.

arXiv:1603.04713

Modeling Time Series Similarity with Siamese Recurrent Networks

Traditional techniques for measuring similarities between time series are based on handcrafted similarity measures, whereas more recent learning-based approaches cannot exploit external supervision. We combine ideas from time-series modeling and metric learning, and study siamese recurrent networks (SRNs) that minimize a classification loss to learn a good similarity measure between time series. Specifically, our approach learns a vectorial representation for each time series in such a way that similar time series are modeled by similar representations, and dissimilar time series by dissimilar representations. Because it is a similarity prediction models, SRNs are particularly well-suited to challenging scenarios such as signature recognition, in which each person is a separate class and very few examples per class are available. We demonstrate the potential merits of SRNs in within-domain and out-of-domain classification experiments and in one-shot learning experiments on tasks such as signature, voice, and sign language recognition.

arXiv:1602.09065

Evaluation of Deep Learning based Pose Estimation for Sign Language Recognition

Human body pose estimation and hand detection are two important tasks for systems that perform computer vision-based sign language recognition(SLR). However, both tasks are challenging, especially when the input is color videos, with no depth information. Many algorithms have been proposed in the literature for these tasks, and some of the most successful recent algorithms are based on deep learning. In this paper, we introduce a dataset for human pose estimation for SLR domain. We evaluate the performance of two deep learning based pose estimation methods, by performing user-independent experiments on our dataset. We also perform transfer learning, and we obtain results that demonstrate that transfer learning can improve pose estimation accuracy. The dataset and results from these methods can create a useful baseline for future works.

arXiv:1607.06356

Reasoning about Body-Parts Relations for Sign Language Recognition

Over the years, hand gesture recognition has been mostly addressed considering hand trajectories in isolation. However, in most sign languages, hand gestures are defined on a particular context (body region). We propose a pipeline to perform sign language recognition which models hand movements in the context of other parts of the body captured in the 3D space using the MS Kinect sensor. In addition, we perform sign recognition based on the different hand postures that occur during a sign. Our experiments show that considering different body parts brings improved performance when compared to other methods which only consider global hand trajectories. Finally, we demonstrate that the combination of hand postures features with hand gestures features helps to improve the prediction of a given sign.

arXiv:1610.07995

Spatial Relationship Based Features for Indian Sign Language Recognition

In this paper, the task of recognizing signs made by hearing impaired people at sentence level has been addressed. A novel method of extracting spatial features to capture hand movements of a signer has been proposed. Frames of a given video of a sign are preprocessed to extract face and hand components of a signer. The local centroids of the extracted components along with the global centroid are exploited to extract spatial features. The concept of interval valued type symbolic data has been explored to capture variations in the same sign made by the different signers at different instances of time. A suitable symbolic similarity measure is studied to establish matching between test and reference signs and a simple nearest neighbour classifier is used to recognize an unknown sign as one among the known signs by specifying a desired level of threshold. An extensive experimentation is conducted on a considerably large database of signs created by us during the course of research work in order to evaluate the performance of the proposed system

arXiv:1701.01875

Sign Language Recognition Using Temporal Classification

Devices like the Myo armband available in the market today enable us to collect data about the position of a user's hands and fingers over time. We can use these technologies for sign language translation since each sign is roughly a combination of gestures across time. In this work, we utilize a dataset collected by a group at the University of South Wales, which contains parameters, such as hand position, hand rotation, and finger bend, for 95 unique signs. For each input stream representing a sign, we predict which sign class this stream falls into. We begin by implementing baseline SVM and logistic regression models, which perform reasonably well on high quality data. Lower quality data requires a more sophisticated approach, so we explore different methods in temporal classification, including long short term memory architectures and sequential pattern mining methods.

arXiv:1705.01389공식 코드 · Papers with Code

Learning to Estimate 3D Hand Pose from Single RGB Images

Low-cost consumer depth cameras and deep learning have enabled reasonable 3D hand pose estimation from single depth images. In this paper, we present an approach that estimates 3D hand pose from regular RGB images. This task has far more ambiguities due to the missing depth information. To this end, we propose a deep network that learns a network-implicit 3D articulation prior. Together with detected keypoints in the images, this network yields good estimates of the 3D pose. We introduce a large scale 3D hand pose dataset based on synthetic hand models for training the involved networks. Experiments on a variety of test sets, including one on sign language recognition, demonstrate the feasibility of 3D hand pose estimation on single color images.

arXiv:1801.10111

Video-based Sign Language Recognition without Temporal Segmentation

Millions of hearing impaired people around the world routinely use some variants of sign languages to communicate, thus the automatic translation of a sign language is meaningful and important. Currently, there are two sub-problems in Sign Language Recognition (SLR), i.e., isolated SLR that recognizes word by word and continuous SLR that translates entire sentences. Existing continuous SLR methods typically utilize isolated SLRs as building blocks, with an extra layer of preprocessing (temporal segmentation) and another layer of post-processing (sentence synthesis). Unfortunately, temporal segmentation itself is non-trivial and inevitably propagates errors into subsequent steps. Worse still, isolated SLR methods typically require strenuous labeling of each word separately in a sentence, severely limiting the amount of attainable training data. To address these challenges, we propose a novel continuous sign recognition framework, the Hierarchical Attention Network with Latent Space (LS-HAN), which eliminates the preprocessing of temporal segmentation. The proposed LS-HAN consists of three components: a two-stream Convolutional Neural Network (CNN) for video feature representation generation, a Latent Space (LS) for semantic gap bridging, and a Hierarchical Attention Network (HAN) for latent space based recognition. Experiments are carried out on two large scale datasets. Experimental results demonstrate the effectiveness of the proposed framework.

arXiv:1805.06618

Optimization of Transfer Learning for Sign Language Recognition Targeting Mobile Platform

The target of this research is to experiment, iterate and recommend a system that is successful in recognition of American Sign Language (ASL). It is a challenging as well as an interesting problem that if solved will bring a leap in social and technological aspects alike. In this paper, we propose a real-time recognizer of ASL based on a mobile platform, so that it will have more accessibility and provides an ease of use. The technique implemented is Transfer Learning of new data of Hand gestures for alphabets in ASL to be modelled on various pre-trained high- end models and optimize the best model to run on a mobile platform considering the various limitations of the same during optimization. The data used consists of 27,455 images of 24 alphabets of ASL. The optimized model when ran over a memory-efficient mobile application, provides an accuracy of 95.03% of accurate recognition with an average recognition time of 2.42 seconds. This method ensures considerable discrimination in accuracy and recognition time than the previous research.

arXiv:2610.00881v1cs.CV

수어를 위한 기계 번역

Machine Translation for Sign Languages

수어 기계 번역(machine translation)은 지난 10년간 고립 수어 인식에서 종단간(end-to-end) 번역 시스템으로 발전하며 상당한 진전을 이루어 왔다. 포즈 추정, 트랜스포머(transformer) 아키텍처, 대규모 데이터셋(dataset) 수집의 발전이 이러한 진전을 이끌었지만, 여전히 과제가 남아 있다. 데이터셋은 음성 언어 자원에 비해 제한적이며, 평가 지표는 출력물의 언어적 품질을 충분히 포착하지 못하고, 모델은 수어의 동시적이고 다층적이며 3차원적인 구조를 포착해야 한다. 본 논문은 기술적 과제와 이해관계자 고려사항 간의 균형을 추구하는 포괄적인 리뷰를 제공한다. 우리는 수어를 계산적으로 독특하게 만드는 언어학적 특성을 살펴보고, 인식, 번역, 생성 시스템의 발전 과정을 추적하며, 현재진행 중인 기술적 과제를 분석한다. 결정적으로, 우리는 데이터 거버넌스, 커뮤니티 참여, 적절한 사용에 관한 윤리적 고려사항을 다룬다. 컴퓨터 비전, 수어 언어학, 농인학(deaf studies)을 아우르는 학제간 관점을 바탕으로, 우리의 분석은 지속적인 진전을 위해서는 이러한 분야들 간, 그리고 농인 커뮤니티와의 지속적인 협력이 필요함을 강조한다.

arXiv:2609.25862v1cs.CL

아이슬란드 수어 고립 수어 인식: 저자원 환경에서의 실험

Isolated Sign Language Recognition for Icelandic Sign Language: Experiments in a Low-resource Setting

본 논문은 아이슬란드 수어(\'ITM)에 대한 고립 수어 인식(ISLR) 최초의 실험을 제시한다. 아이슬란드어–\'ITM 이중 언어 온라인 사전에서 파생된 데이터셋인 \'ITM SignWiki를 사용한다. 이 데이터셋은 실제로 저자원 환경에 해당한다. 1,845개의 비디오가 849개 클래스를 이루며, 이 중 86%는 예시가 단 두 개뿐이어서 전체 과제는 사실상 서로 다른 수어 사용자 간의 원샷(one-shot) 인식이 된다. 어휘 크기가 점차 커지는 세 가지 과제(22개, 117개, 849개 클래스)에서 두 가지 오픈소스 ISLR 프레임워크인 OpenHands와 SPOTER를 비교하고, 세 가지 포즈 추정기와 두 가지 형태의 교차 언어 전이(cross-lingual transfer)를 평가한다. \'ITM 데이터만 사용할 경우 세 과제 모두에서 SPOTER가 OpenHands보다 우수한 성능을 보였으며, MediaPipe 포즈가 AlphaPose나 SDPose보다 더 좋은 결과를 냈다. 가장 큰 성능 향상은 교차 언어 전이에서 나타났다. \'ITM으로 파인튜닝하기 전에 미국 수어 데이터로 SPOTER를 사전학습하면 정확도가 14~24%p 상승하여 세 과제에서 각각 72.7%, 47.9%, 22.6%에 도달했으며, 다른 여섯 개 수어의 데이터를 활용한 다국어 학습은 전체 과제에서 OpenHands의 정확도를 1.41%에서 28.86%로 끌어올렸다. 실용화 수준에는 크게 못 미치지만, 이 결과는 자원이 더 풍부한 수어로부터의 전이가 극저자원 수어에 유망함을 시사한다. 두 프레임워크를 각각 수정한 버전을 공개한다.

arXiv:2607.11341v1cs.CL

The In-Car Sign Language Corpus (ICSL): A Multi-Modal Resource for Constrained-Space Sign Language Recognition

This paper addresses the challenges of using sign language within shared mobility services, such as taxis, carpools, or ride-sharing platforms. The use of sign language recognition (SLR) in real-world, confined environments, specifically vehicle interiors remains largely unexplored. To motivate research in this area, we present the In-Car Sign Language (ICSL) dataset for Brazilian Sign Language (Libras), with the long-term goal of improving public transport accessibility for the Deaf and Hard-of-Hearing community. The dataset consists of: (1) high-precision laboratory motion capture (MoCap) data to establish an idealized linguistic baseline and (2) real-world multi-modal in-car recordings captured using a 2D camera and 3D Time-of-Flight sensors. The dataset provides a basis for comparative analyses between synthesized signing avatar animations and recorded real signing interpreter videos, which enable future research into robust "in-the-wild" SLR models and domain adaptation. We describe in detail the use cases, the setup, the data collection protocol, and the metadata structure of the corpus. In total, we recorded a multimodal dataset exceeding 1.5 million frames, comprising the synchronized multimodal streams described above featuring Libras users across various in-car scenarios. The corpus is provided with gloss annotation of lexical signs and non-lexical sign language elements specially designed to support the training and evaluation of deep neural networks for constrained space recognition. In-vehicle signing offers a technically significant example of a constrained, occluded, and non-frontal environment. While recognizing the diverse communication strategies already employed by the Deaf community, identifying automotive-specific limitations provides a useful stepping stone for research into enhancing in-car accessibility and passenger quality of life.

arXiv:2607.09890v1cs.CV

트랜스포머 시간적 헤드와 풀링 후 모션 게이트가 CorrNet 기반 CSLR에 도움이 되는가? 실증적 연구

Do Transformer Temporal Heads and Post-Pooling Motion Gates Help CorrNet-based CSLR? An Empirical Study

CorrNet은 시각적 인코딩 단계 내에서 프레임 간 상관관계를 모델링하기 때문에 연속 수어 인식(CSLR)의 강력한 기준선(baseline)이다. 본 논문에서는 재현된 CorrNet 시스템의 두 가지 자연스러운 확장을 연구한다: BiLSTM 시간적 헤드를 트랜스포머(Transformer) 인코더로 교체하는 것과, 시간적 풀링 이후에 모션 단서를 주입하는 것이다. 트랜스포머 헤드는 트랜스포머에 맞게 조정된 학습 전략을 사용하더라도 BiLSTM 기준선을 능가하지 못하며, 두 헤드의 계산 및 런타임 비용은 거의 동일한 것으로 나타났다. 두 번째 확장으로, 우리는 MotionGate라는 경량 모듈을 설계하였다. 실험에서 MotionGate는 일관되게 항등 매핑과 유사한 형태로 붕괴한다: 게이트는 모션 선택성을 잃고, 주입된 잔차(residual)는 풀링된 특징에 대한 약하고 비선택적인 섭동이 된다. 이러한 결과는 CorrNet의 상관관계 기반 인코딩 이후의 명시적 모션 주입이 대체로 불필요하며, CSLR에서 자연스러워 보이는 아키텍처 확장은 도움이 된다고 가정하지 말고 신중하게 검증해야 함을 시사한다.

arXiv:2606.28667v1cs.CL

수어 모델의 음운 지각

Phonological Perception of Sign Language Models

수어(Sign Language)는 손 모양(handshape), 위치(location), 동작(movement)과 같은 어휘 하위 음운 매개변수를 조합하여 의미를 형성하는 합성 체계이다. 수어 인식(SLR, Sign Language Recognition)을 위한 심층 학습(deep learning) 모델이 번역 벤치마크에서 성능 향상을 이루었지만, 이러한 모델들이 추상적인 음운 특징을 구별하는지, 아니면 단순히 저수준 통계적 상관관계에 의존하는지는 여전히 불분명하다. 본 연구는 미국 수어(ASL)로 학습된 SLR 모델의 음운 지각을 평가하기 위해 최소 대립쌍(minimal pairs)을 활용한 음운 민감도 탐침(probing)과 인간 행동 데이터와의 표현 정렬(representational alignment) 평가를 수행한다. 실험 결과, SLR 모델은 창발적(emergent) 음운 민감도를 보이지만 명확한 아키텍처 간 트레이드오프가 존재함을 밝혔다: 포즈 기반 모델은 손 모양 대조에 민감한 반면, 픽셀 기반 모델은 위치 변화를 더 잘 포착한다. 또한 포즈 기반 모델은 인간의 지각적 유사도 판단과 상관관계를 보이는 잠재 표현을 학습한다(r~0.49). 이러한 결과는 SLR 모델이 창발적 음운론을 보이지만, 현재의 학습 패러다임이 아키텍처의 귀납적 편향(inductive bias)을 넘어 스케일링(scaling)하기에는 불충분함을 시사한다.