Papers by Vasily Moshnyaga
ECTI Transactions on Computer and Information Technology, 1970
대한전자공학회 ISOCC, Oct 1, 2004
A number of techniques to reduce cache leakage have so far been proposed. However, it is not clea... more A number of techniques to reduce cache leakage have so far been proposed. However, it is not clear that 1) what kind of algorithm can be considered and 2) how much they have impact on energy and performance. To answer the questions, this paper classifies cache-leakage reduction techniques and evaluates their energy-performance efficiency. As a result, we have found that an approach employed by the Drowsy cache [1] achieves the best energy-performance efficiency with low complexity. Moreover, we investigate the potential of the approach on multi-thread program executions.
IEE proceedings, 1997
The authors present a tool for generating a performance-driven placement from a netlist of regist... more The authors present a tool for generating a performance-driven placement from a netlist of register-transfer level (RTL) blocks. Based on the modified force-directed algorithm, the tool chooses locations and orientations of the blocks in such a way as to produce a compact area placement with minimum wiring delay along the critical path. Experiments show that the tool provides solutions close to those generated manually, is fast enough to be used in the inner loop of a program which synthesises RTL structures from behavioural specifications and ensures the strong links between RTL synthesis and timing-driven layout which are so necessary for design of submicron ASICs.
Lecture Notes in Computer Science, 2006
Caregiving a person suffering from dementia or loss of brain cognitive ability due to aging is a ... more Caregiving a person suffering from dementia or loss of brain cognitive ability due to aging is a big physical, mental and emotional burden to family members. In this paper we present a novel system for assisting caregivers at home. The system employs heterogeneous sensing and artificial intelligence technologies to automatically and unobtrusively monitor the patient; assess possible risks that the patient may face in current situation; and alert the caregiver on emergency by delivering video, audio and text to his mobile phone or PC. We discuss the system architecture and technologies applied for sensing, communication, assessment and user-interface, and present a prototype system implementation.
Falls are a major public health concern and the leading cause of accidental deaths among elders. ... more Falls are a major public health concern and the leading cause of accidental deaths among elders. Technologies capable of fast and accurate fall detection are crucial to provide immediate assistance and reduce complications associated with a fall. This article presents a new deep-learning approach for automated fall detection from signatures of a continuous-wave micro-Doppler radar. Unlike related methods, the proposed approach applies convolutional neural networks directly to raw time-frequency spectrograms to identify falls. It allows CNN to learn the fall features and classification boundaries without employing additional hardware or large data sets to distinguish abrupt and slow falls from non-fall actions. Experiments show that the approach achieves the fall classification accuracy of 95%.
Springer eBooks, 2015
People suffering from dementia or loss of brain cognitive ability due to aging can make a judgmen... more People suffering from dementia or loss of brain cognitive ability due to aging can make a judgmental error and therefore have to be constantly monitored. In this paper we present a smart-carpet that employs sensing technology to automatically and unobtrusively monitor a person, detect fall and alert the caregiver by his phone or PC. In contrast to existing solutions our design is effective yet inexpensive, easy to install and maintain. We discuss the proposed design, present its prototype implementation and results of empirical evaluation.
IGI Global eBooks, Nov 25, 2013
With the explosive use of personal computers or PCs, reducing computer energy consumption is para... more With the explosive use of personal computers or PCs, reducing computer energy consumption is paramount for sustainability. The display is the largest energy consumer in a personal computer. Current display energy management technologies ignore the attention of the PC user and therefore may either switch the display off when the user looks at the screen or lose energy by keeping the display on when nobody looks at it. This chapter discusses a new display energy management technology and outlines its implementation in a personal computer system. Unlike existing technologies, which “sense” a PC user through keyboard and/or mouse or the other sensors, this technology “watches” the user through a single camera or CMOS vision sensor. The technology tracks the user’s eyes, keeping display active only if the user looks at its screen. Otherwise, it dims the display down or even switches it off to save energy. The authors implemented the technology in software and hardware and present the results of their experimental evaluation.
International Conference on VLSI and CAD, 1993
Reducing energy consumption has become an important design problem. Although most of consumer ele... more Reducing energy consumption has become an important design problem. Although most of consumer electronic systems are user centric existing energy management methods are device centric. The methods either assume unchangeable operational environment or rely on very simplified policies, which eventually lead to large energy losses. In this paper we present a new energy management approach that allows systems to monitor their users for energy saving. We discuss applications of this novel approach in personal computers, television sets and home automation system and show the results of their experimental evaluation. I.
IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences, Apr 1, 2002
Asia and South Pacific Design Automation Conference, Jan 21, 2003
Studies in systems, decision and control, 2018
Technology has entered the age of smart systems, which not only implement traditional human funct... more Technology has entered the age of smart systems, which not only implement traditional human functions of analyzing real-world data and making decisions but also employ humans within the feedback control loop. The development of such systems is not trivial, however, involving several challenges. In this chapter, we share our experience of building human-in-the-loop systems, such as a user-aware computer display, a viewer-conscious TV, a smart door, a smart carpet, a medication adherence control system and smart in-home system for monitoring people with cognitive impairment. We identify problems related to incorporating humans into the control loop and present guidelines for hardware and software designers.
2022 IEEE 4th Global Conference on Life Sciences and Technologies (LifeTech)
Falls are a major public health concern among seniors living independently at home. This article ... more Falls are a major public health concern among seniors living independently at home. This article presents a new approach for automated real-time fall detection by a single continuous-wave micro-Doppler radar and deep learning. Unlike related methods, we apply Long Short-Term Memory based recurrent neural network (RNN) to time-frequency spectrograms of Doppler-radar returns to identify falls. The approach does not require extra hardware or big data set to classify falls from non-fall actions and has superior accuracy compared to CNN-based fall detection.
This paper presents an optimized MPEG2 video codec implementation, which drastically reduces the ... more This paper presents an optimized MPEG2 video codec implementation, which drastically reduces the number of computations and memory accesses required for video compression. Unlike traditional scheme, we reuse data stored in frame memory to omit unnecessary coding operations and memory read/writes for unchanged macroblocks. Due to dynamic memory sharing among reference frames, data-driven macroblock characterization and selective macroblock processing, we perform less than 15% of the total operations required by a conventional coder while maintaining high picture quality.
2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS), 2017
This article presents new approach for unobtrusive indoor fall detection by an IR thermal array s... more This article presents new approach for unobtrusive indoor fall detection by an IR thermal array sensor. Unlike existing methods that run fall detection at a server and require high communication rates, we perform fall detection within the sensor node by using a computationally inexpensive algorithm that only signals the server when a fall occurs. Experiments with prototype design show that such formulation provides robust and real-time fall detection even in a noisy environment.
2018 IEEE 61st International Midwest Symposium on Circuits and Systems (MWSCAS), 2018
This article describes a new technique for unobtrusive detection of medication intake in a non-we... more This article describes a new technique for unobtrusive detection of medication intake in a non-wearable system equipped by Kinect. Unlike existing methods that rely on skeletal modeling, the proposed technique fuses the depth image data with the RGB image data, making decisions based on both the spatial coordinates and the mouth features. Experiments show that such formulation provides the medication intake recognition in real time with better accuracy than existing methods.
2018 IEEE International Workshop on Signal Processing Systems (SiPS), 2018
Fingerprints is nowadays the most largely used biometric technology for personal identification. ... more Fingerprints is nowadays the most largely used biometric technology for personal identification. With emerging ubiquity of high-definition digital cameras, the risk of fingerprint hacking from photo and video pictures increase. This article discusses the fingerprint hacking problem and presents a new camera-based anti-hacking approach. The technique automatically scans a captured picture for fingerprint images and if found modifies the images to protect the fingerprint information. An experimental evaluation of software prototype shows that it provides robust detection of fingerprints with less than 1sec. delay for 1600x1200 pixel images and almost invisible impact on image quality (less than 1% increase in PSNR).
2017 IEEE International Conference on Consumer Electronics (ICCE), 2017
Muslim praying ritual is quite complex and challenging especially for old and forgetful people. T... more Muslim praying ritual is quite complex and challenging especially for old and forgetful people. This paper presents new smart mat system to assist Muslims in praying. The proposed system computes starting time of a pray, reminds the user about the pray time, indicates direction to Mecca and counts the number of postures performed during a pray. Unlike related formulations, it does not require the user to possess smartphones or other devices and utilizes new approach for posture detection. The electronic parts embedded into the mat are inexpensive yet efficient. Experiments with the smart mat prototype show that it has good accuracy, easy to use and is helpful for Muslim prayers.
TENCON 2015 - 2015 IEEE Region 10 Conference, 2015
Reversible digital multipliers theoretically do not dissipate energy but are very complex in desi... more Reversible digital multipliers theoretically do not dissipate energy but are very complex in design. To reduce the gate count, many proposed optimizations employ large reversible gates, which however neither decrease the circuit complexity nor its quantum cost. In this paper, we present design techniques that result in minimal complexity of 4×4-bits unsigned reversible multiplier in terms of the number of gates, the number of garbage outputs, the number of constant inputs and the total quantum cost.
Uploads
Papers by Vasily Moshnyaga