Minsik Kim

Accelerating HEVC transcoder by exploiting decoded quadtree

This paper proposes an accelerated High-Efficiency Video Coding (HEVC) transcoder which can promptly provide downscaled video contents to various devices. The quadtree information is first extracted from the decoding process and is transformed for the target resolution. By utilizing the decoded depth information, the encoder can be accelerated by only searching the optimal depths in the quadtree, without losing the video quality. The encoder also adaptively changes the depth search ranges according to the picture order count (POC) for further optimization. Our proposed method shows a maximum encoding speedup of 2.18, with only 0.3% BD-rate increase in the best case.

A Light guide plate based flexible optical cuff for optogenetic stimulation of motor units

This paper reports a light guide plate based flexible optical cuff for optogenetic stimulation of motor units. We propose the optical cuff based on a single-sheet PDMS light guide plate (LGP). It has good flexibility, and can be improved the efficiency of the light source by distributing the light globally through the LGP. Moreover, the possibility of cell necrosis due to heat can help to reduce by preventing direct contact between the μ-LED and neurons.

CANYVAL-X Mission Development Using CubeSats

Current space telescopes have a single structure, and consequently, their focal length cannot be increased sufficiently. Sometimes, this problem may prevent the improvement of the resolution of the telescope. To solve this problem, the concept of virtual telescope has been proposed. A virtual telescope consists of two spacecraft: one has a lens system and the other has a detector system. By using formation flying, the two spacecraft can be simplified as a virtual telescope system. Then, their relative orbit distance can serve as a baseline for the virtual telescope system [1, 2]. The most important issue in a virtual telescope is to perform inertial alignment with respect to a celestial object and to maintain it in space. Inertial alignment means that the relative position and relative attitude of the two spacecraft are simultaneously aligned with a target.

An accelerated separable median filter with sorting networks.

Since the computer vision technologies are used to extract features from the edges of images, the median filter is frequently used as a pre-processing operation to improve the quality of feature objects. However, the median filter algorithms still have the limitation on integrating with the mobile systems due to the large computation overhead. Since the median filter accounts for a large proportion of total computation, this paper proposes a separable median filter to resolve the limitation. The proposed separable median filter is implemented with the sorting network, and reuses the data sorted at neighbor pixels, reducing duplicated comparisons.

Accelerating gesture recognition algorithm using coarse grained reconfigurable architectures

The gesture recognition algorithms have been widely used to realize the human-computer interaction in multimedia applications, but still have limitation of integration due to the computation overhead. For this reason, this paper proposes an acceleration method for the hand gesture recognition algorithm using CGRA. The proposed algorithm enables software pipelining and vectorization for the SRP architecture, which exploits both the instruction level parallelism and data level parallelism. Therefore, the proposed optimization method improves the utilization of the hardware resources in SRP and effectively accelerates the performance of the gesture recognition algorithm, achieving maximum speedup of 10.97.