Search

This paper proposes a system that can be installed on a shopping cart that be used to assist visually impaired people. This method uses the concept of ROI, text localization and recognition by optical character recognition software and a webcam controlled by Raspberry Pi. Two micro controllers are used to program both the shopping cart movement and OCR. The user is required to give a voice input of the required product and the ROBOKART guides the user to that product.

We present an end-to-end system for creating talking head video from text and a single image automatically. A two-dimensional clone of an individual made entirely of a still image of a person, that conveys the given text message to the target audience through replicated facial expressions. It can send any message that is delivered in a text to it with a voice that is cloned from a few seconds of their actual voice input. Our technique yields videos of natural facial movements including blinks and brow movements, as well as mouth movements that are in synchronization with the audio. The proposed method only uses a single base model to create a personalized video, avoiding the hassle of training it every time.

An assistive training tool for rehabilitation of dysphonia is designed, developed and iteratively evaluated and optimized according to the practical clinical needs. The assistive tool employs serious games as the attractive logic part, and running on the tablet with normal microphone and/or Kinect2 as voice input device. The patient is able to play the game as well as conduct the voice training simultaneously guided by therapists at clinic, while not be interfered, or do rehabilitation exercise independently at home. The voice information can be recorded and extracted for evaluation of the long-time rehabilitation progress. This paper outlines a design science approach for the development of an useful software prototype of such a tool. Both therapists and sufferers have provided earnest suggestions for the improvement. Seven pitch estimation algorithms have been evaluated and compared with selected patients voice database. A series of benchmarks have been generated during the evaluation process for technology selection. Finally, a refined emotion wheel questionnaire is designed to evaluate the controllability and the happiness.

We designed and implemented cross-cultural learning activities for this study. Participants from two countries, geographically located very far from each other, represented different cultures that have no communication language in common. Two systems were applied to learning activities in order to enable interaction and information exchange among the participants: (1) a speech-to-text recognition system, which generates texts from a speaker's voice input in his/her native language and (2) computer-aided translation system, which simultaneously translates texts into the language of the speaker's foreign peers. The goal was to test the feasibility of learning activities supported by the two systems and their effectiveness for cross-cultural learning. To this end, we evaluated participants' learning outcomes, analyzed their online communication with peers, and carried out a questionnaire survey and interviews with both the participants and their instructors. The use of multiple data sources allowed triangulation of the findings, thus adding rigor to the research. We obtained three findings through this study. First, cross-cultural learning took place. Second, the questionnaire and interview results show that the two systems are easy to use and useful for cross-cultural learning. According to participants, even with no common language, they could still interact and exchange culture-related information using the two systems. Finally, the results indicate that the texts produced by the two systems are acceptable and useful for the cross-cultural learning of participants (except texts translated from Russian into Chinese in the second week). These findings suggest that the educators and researcher can implement cross-cultural learning activities for participants with no common language with the support of speech-to-text recognition and computer-aided translation systems, as these two systems can help participants to communicate and exchange culture-related information. Our approach enables cross-cultural learning.The two systems are easy to use and useful for cross-cultural learning.The texts produced by the two systems are acceptable and useful for learning.

With the advancements in the hardware industry, an increase in the computation power and development in Artificial Methods, we can think of working on cognitive tasks. We have worked on various speech recognition methods using Natural language processing and Hidden Markov Models. We have done the classification of the users on the basis of their utterances. In this paper, we propose a discrete probability approach. The results which we have got gives us high accuracy results in recognizing the speakers. This helps in concluding that Learning and Uncertain reasoning are important components of Artificial Intelligence that could help in the development of solutions to problems which are interesting and complex.

Previous studies confirmed the correlation between dream and creativity. If we can make good use of our dreams, they may bring our big breakthrough through non-logic thinking state of brain. This study is aim to develop a dream journal or dream recording device that for recording ideas in dreams. Through analyzing the participants’ writing habits of dream reports to define the criteria of dream journal design. First this study implemented one week dream report experiment to observe the recording behaviors of participants. In the focus group method, this study collected about 77 dream journal design key words from both the interview and the participants’ advices. After that, using KJ technique to organize the key words and come out 5 design criteria in result. Base on the criteria and the result that discussed with the participants, this study presented a dream journal prototype. It uses the voice to record users’ creative dreams. The sensor on pillow controls the power of the dream journal device and adjusts its projection. Users make the dream image board with voice input and edit contents with gestures. After finishing recording dreams, the dream journal device shows dream image lights. Users can print out the dream journal or share them to others. If users run out of inspirations, they can take a look at dream journal to help them come out with some new ideas.

Voice based human-computer interactions have raised much interest and found various applications. Some extant voice based interactions only support voice commands with fixed vocabularies or preset expressions. This paper is motivated to investigate an approach to implement a flexible voice-enabled CAD system, where users are no longer constrained by predefined commands. Designers can, to a much more flexible degree, communicate with CAD modelers using natural language conversations. To accomplish this, a knowledge-guided approach is proposed to infer the semantics of voice input. The semantic inference is formulated as a template matching problem, where the semantic units parsed from voice input are the ''samples'' to be inspected and the semantic units in the predefined library are the feature templates. The proposed behavioral glosses, together with CAD-specific synonyms, hyponyms and hypernyms are extensively used in the parsing of semantic units and the subsequent template matching. Using such sources of knowledge, all the semantically equivalent expressions can be mapped to the same command set, and the Voice-enabled Computer Aided Design (VeCAD) system is then capable of processing new expressions it has never encountered and inferring/understanding the semantics at runtime. Experiments show that this knowledge-guided approach is helpful to enhance the robustness of semantic inference and can effectively eliminate the chance of overestimations and underestimations in design intent interpretation.

In this paper, some of the results for the development of an artificial brain for LifeRobot ''Tarou'' developed at Artificial Life and Robotics Laboratory of The University of Oita are introduced. Solutions for questions on how to build the artificial brain and how to implement the artificial brain are presented from vision and voice input signals to wheel control, head control, camera control, speech, touch screen, LED status output. Also, the research issues of Tarou are stated and discussed. It is shown that the artificial brain developed provides Tarou with abilities to perceive an environment, to interact with humans, to make intelligent decisions and to learn new skills.

The increasing power of computer hardware is enabling the user interface to enter a new dimension, where voice enhanced interfaces can remove the barrier between people and the machine for some users. In this study the traditional keyboard and mouse interface (TI) was compared with voice enhanced interface (VEI). An automatic semester grade calculation software was developed in visual basic and enhanced with a “trainable” voice recognition software for Cantonese and English. Results show that there are significant differences in task completion time, number of errors and satisfaction for the traditional keyboard interface and the voice enhanced interface. In addition, contrary to our intuition, native Cantonese speakers had more voice input errors when speaking Cantonese than English. Though voice enhanced interfaces have many potential applications, consideration should be given to the differences between English and spoken languages that are based on tones such as Cantonese. It appears that further development in voice recognition technology is required in order to make voice input widely usable for the Chinese speaking computer users. Relevance to industry Voice input may seem the ultimate input method, since hands free data entry can allow tremendous flexibility. However, the results of this study indicate that, in industry, voice input should be used with caution, especially when tonal languages or safety are involved.

This paper presents an investigation into computer tool support for creative tasks in the visual arts, in particular sketching. The research draws from both theoretical models of creative processes and empirical observations of sketching activity in the domain of character development in animation and illustration. These sources show that creative problem solving through sketching is a highly iterative process during which various constraints are rapidly evolving and being adapted to. A set of requirements for an experimental, multimodal sketching tool was developed, drawing from our theoretical and empirical studies. In order to support the intense, rapid nature of the task, voice input was implemented allowing the user to access functionality without interrupting the hand activity. The multimodal tool also included radial marking menus in order to provide rapid navigation to functionality. Informal evaluations were conducted that provided qualitative feedback about the use of the system.

The objective of this research was to evaluate voice enhanced interface (VEI) and traditional keyboard and mouse interface (TI) in terms of users' performance, degree of satisfaction and mental workload. A trainable voice recognition software for student grade calculation was developed for the study, tested for Mandarin and English voice input, and compared to traditional input. The results show significant differences in task completion time, number of errors, satisfaction and mental workload for VEI and TI. The subjects were more satisfied with keyboard and mouse input. However, the attitude toward voice enhanced interface may change with time if improved recognition is achieved. A discussion about the comparison between the Mandarin and Cantonese voice input is also included in this paper.

This paper describes the essential speech processing techniques for interactive voice applications in the telecommunications field. These techniques include speech recognition and speech synthesis, both of which aim to make interactive speech communications between man and machine more natural. Keyword spotting, background noise effects reduction, and speakers and/or telephone adaptation techniques are considered essential in speech recognition in order to allow a more natural voice input as well as an adequate robustness against environmental variabilities. In the area of text-to-speech synthesis, we propose a rule-based synthesis method applicable to the Japanese language, aiming to produce high quality speech. The commercial system ANSER of a former project is also described as an example of an interactive speech processing system. Finally, a recently developed speech recognition server which includes a vocabulary-flexible recognition function is described. It is meant to illustrate the concept of the techniques it employs which allow its range of applications to be easily extended and also allow it to adapt itselt to the changes which are rapidly occurring in the telecommunications field.

This report describes two procedures that aim to evaluate image quality. The first is an implementation of the ISO performance-based for visual displays. Subjects are presented with random characters presented in five blocks on the screen and asked to identify each character. This test evaluates displays on three dependent measures: average times, error rates and comfort ratings by subject. When the error rate and the subjective ratings are considered, the results that the adequately discriminates a display that meets the physical requirements of the standard from one toes not. However, when full keyboard input is used, timing scores are badly affected by practice effects. In a second modified procedure, subjects work for a fixed period of time and the task is a four-alternative forced-choice procedure with catch trials. This new procedure can distinguish a display that passes the physical measurements of part 3 of ISO 9241 from one that fails, on all three dependent measures. Compared with the original performance test method proposed in part 3 of ISO 9241, the new test method has a number of advantages. The results from this study suggest that the new test procedure is: (1) Valid: all dependent measures concur; (2) Simple: subjects learn the task quickly, and it is easy to analyse the data in real time; (3) Efficient: more than three times as fast as the old test; and (4) Flexible: keyboard, mouse or voice input may be used.

The U.S. livestock and meat industries depend heavily on the USDA to provide the rating of their meat for marketing. Since the carcass-beef grading system was put in place in 1927, carcasses have been graded by human graders. To achieve better consistency and equity, a knowledge-based expert system for assisting graders to grade carcass beef has been developed. This paper presents the development of the knowledge-based expert system and discusses how the expert system was integrated with a voice-recognition subsystem that provides the characteristics of the carcasses to the expert system.

An integrated system of hardware and software has been developed to combine the input of coordinate data from a digitizing pad with voice input for object identification or classification. The menu and voice-controlled software generates a sequential ASCII file which contains an object identification section and a data section. Rules for data entry and analysis programs have been developed for several applications. These are being tested and include general stereological analyses, bone, kidney and skin histomorphometry, boundary analyses, neuron classification and malignancy grading.

In conclusion, several computer programs have been developed to aid diagnosis in dermatology. These have used various methods to reach a diagnosis, with some earlier programs using probabilistic techniques and later programs using cognitive models. Several current trends, if they continue, would indicate that computer-assisted diagnosis may play a greater role. The growing number of physicians who own the prerequisite computing resources will probably want diagnostic software to accompany their office management and data base software. The increasing ease of use of diagnostic software, with rapid access and ease of updating, should make these programs more attractive to physicians. Some of the newer peripheral devices such as optical disks and voice input will probably make the software more interesting. Perhaps the most important factor governing the acceptance of computer diagnostic adjuncts is whether the software will provide a useful service to physicians. If these programs can be demonstrated to improve patient care, they may become commonplace in physicians' offices.

Recently, automatic speech recognition systems have shown the potential of becoming a useful means of data entry and control. The most successful of these speech recognition systems accept an isolated utterance as input and use a task-oriented syntactically-constrained vocabulary for increased recognition accuracy. At the same time, language-directed editors are beginning to be introduced into the programmer's workplace. A language-directed editor is an editor that has knowledge of the underlying syntax (and basic semantics) of a language. Program entry, then, is syntax-driven and program editing may proceed on a syntactic (semantic) basis. This article discusses the design, implementation, and results of a controlled experiment to evaluate voice versus keyboard (the standard input mode) in a language-directed editing environment. Twenty-four subjects inputted and edited program segments under control of a language-directed editor via the two input modes. Measures of speed, accuracy, and efficiency were used to compare these two modes of input. In general, the results showed that the subjects were able to complete more of the input and edit tasks by keyboard (70%) than by voice (50–55%), but that keyboard input had a higher error rate than did voice input. Also, the use of voice was just as efficient as keyboard for the inputting of editing commands. These results must be viewed with the understanding that the subjects were novices with respect to voice input, but were very experienced with keyboard input. In this light, it can be seen that voice holds much promise as mode of input for computer programming.

Today's voice technology can provide voice input/output systems that can be used effectively in the factory. This will happen only if the proper technology is connected to suitable applications. Discrete word-dependent speaker recognition in inspection data entry is one such combination.

Summary The paper provides a new method for the application of the speech recognition and analysing technigue to CNC machine tools and other machines. The speech is firstly normalized in amplitude and nonlinearly normalized in time by the combination of frequency analysis and time analysis. According to the specific grammar, the system hierarchically scans, processes and recognizes input speech by using “eight-grade” method. The system also has functions of graph display, machining trail tracking, code transformation and self testing.

The traditional office relies heavily on the use of letters, memos and the telephone for human interaction. Written communications are slow and difficult to manage. Some societies predominantly use the telephone, with the disadvantages of expense and unrecorded business transactions. Now that multinational, global conglomerates dominate world business, the demand for effective, rapid, recorded communication between managers has increased. The author describes her experiences as a global communicator at AT&T International Japan, where she is senior marketing manager, responsible for introducing the AT&T data network and computer product line into Japan. AT&T Japan has installed an integrated office system comprising a small business computer or terminal for each manager, a building-wide LAN and a digital PBX. All managers can rapidly communicate by electronic mail with each other, with customers in Japan, and with the large number of AT&T support and headquarters organizations in the USA and the rest of the world. Previously the subsidiary communicated by facsimile and the telephone after midnight. Marked productivity improvements have resulted from the new methods, and the coordination of a major international development project is outlined. The paper also describes the advanced Total Office Automation System already in use at the Toshiba Headquarters Building in Shibaura, Tokyo. Extensive research led to an integrated system featuring over 1000 units of networked OA equipment, fiber-optic LANs and state-of-the-art distributed data processing, tailored to individual management functions. This hierarchical approach makes clerical work more efficient, aids information processing for middle managers, and helps decision-making by top executives. Finally, future trends such as on-line translation, voice input and open systems networking are discussed, and predictions made about their impact on the global office.

The potential cost benefit from integrating data and voice in a single network provides a strong motivation for packetized voice transmission. In this study, simulation models are used to characterize the delay distribution of voice packets in a single hop as well as in a multi-hop network environment. The trade-off between the number of speakers that can be multiplexed using speech activity detection technique and the delay performance are quantified. This is contrasted with the performance obtained by using a lower bit-rate voice input without speech activity detection. Finally, a procedure is presented which guarantees that the delay for any packet belonging to a call will not exceed a given value throughout the duration of the call, once the call is set up. This procedure can be used with or without speech activity detection.

Present voice input systems have a recognition rate less than 100%. Beside the quality of the voice recognition algorithms some ergonomie factors relating to the speaker, the vocabulary and the environmental noise determine the recognition rate. In a series of experiments 7 parameters (speaker training, long-term speech consistency, system training, speaker sex, vocabulary phonetics, -size, background noise) were tested in a simple word input task. The experiments were performed with different vocabularies (mostly the ICAO alphabet) and with one or both of the voice input systems Threshold T600 and SpeechLab 20A. The results are reported and discussed in the context of the literature.

Speech technology, an amalgam of speech sciences and microelectronics technology, offers new potential to aid the disabled. Low-cost very-large-scale integration (VLSI) speech synthesizers have been on the market for some time and, consequently, the use of synthetic speech output devices in aids has received a great deal of attention. Only recently, however, have VLSI chip-sets for speech recognition started to appear. This article reviews the possibilities for exploiting low-cost speech recognizers in aids for the severely physically disabled. Five areas of application are highlighted: wheelchair control, control of the domestic environment, text composition and editing, and control of manipulators are four of these. The ubiquitous nature of microprocessors and microcomputers, however, makes it difficult nowadays to distinguish sensibly between a general-purpose computer and a computer-based aid. A further important application, therefore, is likely to be provision of a voice interface between a disabled user and a “personal” (or remote) computer, irrespective of the function or functions implemented on that machine. As a consequence, some or all of the above applications could be combined within the same device. It seems sensible, however, at this stage to concentrate on a single application. It is argued that control of the domestic environment by voice is likely to be the most fruitful area for the initial application of automatic speech recognition technology to aid the disabled. The development of a speech interface to an environmental control (ECU) is described, and the principles on which the system design (hardware, software and man-machine dialogue) is based are explained and justified. Evaluation of the system has demonstrated the feasibility of employing speech recognition in environmental control. This is a continuing development, and many possibilities for future improvement are apparent.

The article describes the state of voice recognition and discusses the implementation of a ‘hybrid’ voice input system combining voice input with a traditional means of data input, e.g. a keyboard, in a system for the disabled.