Philippe Mabilleau

16 kbps wideband speech coding technique based on algebraic CELP

The application of algebraic code excited linear prediction (ACELP) coding to wideband speech is presented. An algebraic codebook with a 20 bit address can be used without any storage requirements and, more importantly, with a very efficient search procedure which allows for real-time implementation. The authors describe an efficient procedure for searching such a large codebook deploying a focused search strategy, where less than 0.1% of the codebook is searched with performance very close to that of a full search. High-quality speech at a bit rate of 13 kbps was obtained.<<ETX>>

Touching and hearing GUI's: design issues for the PC-Access system

PC-Access is a system which combines both hardware and software in order to provide multimodal feedback in a Microsoft Windows graphical interface and within its applications. We propose two versions of PC-Access: one which offers sound feedback with an enhanced drawing tablet and another in which tactile stimuli are synthesized by a haptic pointing device. When using the second version, the user will be able to perceive the interface objects (e.g, icons, menus, windows) as well as actions (e.g, moving, re-sizing). Thus, PC-Access offers auditory information (non-verbal sounds and voice synthesis), reinforced by the sense of touch which in turn helps to direct manipulation.

The 'stretched'-Golay and other codes for high-SNR finite-delay quantization of the Gaussian source at 1/2 bit per sample

Error-correcting codes are considered as codebooks for high-performance vector quantization (VQ) of the IID Gaussian source at fractional bit rate. A family of good rate-one-half codes is introduced: the stretched Golay codes. The performance of these codes is compared to other good block codes, trellis-coded quantization, and other techniques. The stretched Golay codes are shown to outperform previously published results for block lengths 32, 40, 56, and 64. The good performance, together with fast decoding make these techniques attractive for applications such as low-bit-rate coding of speech. >

Location estimation in a smart home: system implementation and evaluation using experimental data

In the context of a constantly increasing aging population with cognitive deficiencies, insuring the autonomy of the elders at home becomes a priority. The DOMUS laboratory is addressing this issue by conceiving a smart home which can both assist people and preserve their quality of life. Obviously, the ability to monitor properly the occupants activities and thus provide the pertinent assistance depends highly on location information inside the smart home. This paper proposes a solution to localize the occupant thanks to Bayesian filtering and a set of anonymous sensors disseminated throughout the house. The localization system is designed for a single person inside the house. It could however be used in conjunction with other localization systems in case more people are present. Our solution is functional in real conditions. We conceived an experiment to estimate precisely its accuracy and evaluate its robustness. The experiment consists of a scenario of daily routine meant to maximize the occupants motion in meaningful activities. It was performed by 14 subjects, one subject at a time. The results are satisfactory: the systems accuracy exceeds 85% and is independent of the occupants profile. The system works in real time and behaves well in presence of noise.

Bayesian Filtering and Anonymous Sensors for Localization in a Smart Home

In the context of a constantly increasing aging population with cognitive deficiencies, insuring the autonomy of the elders at home becomes a priority. The DOMUS laboratory aims to address this issue by conceiving a smart home which can both assist people and preserve their quality of life. Obviously, the ability to monitor properly the occupants activities and thus provide the pertinent assistance depends highly on location information inside the smart home. This paper proposes a solution to localize the occupant thanks to Bayesian filtering and a set of anonymous sensors disseminated throughout the house. The localization system is designed for a single person inside the house. It could however be used in conjunction with other localization systems in case more people are present. Our solution is already functional in real conditions. The preliminary results are encouraging: location estimation exceeds 85% certainty in most cases. We conceived an experiment to estimate precisely the systems accuracy. The experiment consists of a scenario of daily routine that maximizes the occupants motion in meaningful activities and will be performed by 30 volunteers. We conclude the paper by discussing possible continuations of this work.

Cooking risk analysis to enhance safety of elderly people in smart kitchen

Risk analysis during cooking activities enables to build a cooking-safe system in order to enhance safety of elderly people in smart kitchen. The Kitchen is the second place where majority of domestic accidents occur, and in particular the oven presents the most principal reason of fire accidents in the residence. The paper presents insightful cooking risk analysis that permits to determine the pertinent parameters to be monitored and measured during cooking in order to prevent risks. We investigated several cooking experiments, and analyzed the composition of heated cooking materials, the concentrations of gases in the cooking smoke, and humidity. The pertinent parameters determined in this paper are: concentrations of gases in the cooking smoke, ambient temperature, utensil temperature, burner temperature, relative humidity, and presence of an object on burner.

Toward Pervasive Computing System to Enhance Safety of Ageing People in Smart Kitchen

Kitchen is the second place where the majority of domestic accidents occur, and in particular oven presents the most principal source of fire accidents in residence. Therefore, enabling kitchen safety is a major factor for ageing people independent living. This paper presents the hardware architecture of our cooking-safe system that targets enhancing safety of ageing people while cooking. The system is based on insightful cooking risk analysis that enables to determine the pertinent parameters to be monitored and measured while cooking. This paper also presents the results of our experimental study that leads us to select the appropriate sensors to constitute the basic building block of our cooking-safe system. The system is composed of sensor nodes to monitor events around oven, then the sensory data is transmitted to a computing unit. The system proactively reacts to hazards in order to prevent cooking associated risks.

Medium band speech coding using a dictionary of waveforms

The paper presents a new approach to the problem of speech coding in the range of 4 to 8 kbits/sec along with experimental tests performed on a MAP 200 array processor. The basic idea consists of using a dictionary of K indexed waveforms of fixed duration T (where T lies between 1 and 4 ms). The input signal is broken down into blocks of duration T. For each segment the Dictionary is searched for that prototype waveform which is, with respect to some distance measure, the closest to the input block. The digital representation of the signal is thereafter the index of the prototype. The paper addresses the problem of the judicious choice of distance measures and the design of a good dictionary. Experimental results are provided for dictionary coding of the signal using a distance measure combining time and frequency domain features. Experimental results are further presented for the coding of the LPC residual.

A comparative study of the proposed high quality coding schemes for digital music

This paper deals with comparisons between various coding schemes for the digital transmission of high quality sound signals typical of what will be transmitted through a Direct Broadcast Satellite System (DBS). Several PCM coding techniques using instantaneous logarithmic coding laws [1] (A-law and µ-law) or near-instantaneous coding law (block coding scheme) such as the NICAM 3 system [2] as well as the Adaptive Delta Modulation (ADM) [3] are compared in this paper. Several pre-emphasis/de-emphasis laws are also proposed to optimally shape the spectrum of the sound program signal before encoding [1]. An investigation of the properties of these laws is carried out and an evaluation, both objective and subjective, of the impact of their utilisation in a codec is presented.

4800 Bps RELP Vocoder using vector quantization for both filter and residual representations

The paper presents the full description and discusses the performances of a 4800 bit per second residual excited linear prediction vocoder. The LPC analysis is efficiently performed using a type of binary-tree search vector-quantization approach. The technique, which is described in ref (1), uses a set of hyperplane equations to perform a hierarchical pattern classification of the input autocorrelation vector in the autocorrelation space. The end result of the search is the integer i 1 which is the index of the most appropriate (in the Itakura-distance sense) prediction filter out of a set of N preset filters. The search requires only \Log_{2}N dot products. In this case vector quantization presents two advantages over the classical approach of the Durbin algorithm followed by scalar quantization. First, a faster algorithm is obtained. Second, the same accuracy in filter representation is possible with less bits per second and consequently more bits can be allocated for representing the residual and gain. The residual is vector quantized in the time domain by blocks of 16 the samples according to the approach of ref (2). The 16 sample block is essentially encoded using the integer I 2 which is the index of the most appropriate 16-sample waveform out of set of M preset prototype waveforms stored in memory. The paper includes preference testings for comparison with other types of 4800 Kbit/sec vocoders. Some sample recordings will be presented at the conference. Finally, preliminary results in the attempt to implement the vocoder in real time on a MAP 200 array processor are discussed.