Evgeny Belyaev

Robust Vehicle-to-Infrastructure Video Transmission for Road Surveillance Applications

IEEE 802.11p vehicle-to-vehicle and vehicle-to-infrastructure communication technology is currently an emerging research topic in both industry and academia. Respective spectrum allocation of 10-MHz channels in the 5.9-GHz band in the United States and Europe allows considering intervehicle transmission of live-video information as a basis, which enables a new class of safety and infotainment automotive applications such as road video surveillance. This paper is first of its kind where such a video transmission system is developed and experimentally validated. We propose a low-complexity unequal packet loss protection and rate control algorithms for scalable video coding based on the 3-D discrete wavelet transform. We show that in comparison with a scalable extension of the H.264/AVC standard, the new codec is less sensitive to packet losses, has less computational complexity, and provides comparable performance in case of unequal packet loss protection. It is specially designed to cope with severe channel fading typical for dynamic vehicular environments and has low complexity, making it a feasible solution for real-time automotive surveillance applications. Extensive measurements obtained in realistic city traffic scenarios demonstrate that good visual quality and continuous playback is possible when the moving vehicle is in the radius of 600 m from the roadside unit.

Live video streaming in IEEE 802.11p vehicular networks : Demonstration of an automotive surveillance application

Prospective IEEE 802.11p-enabled automotive video applications are identified. Preliminary experimental results of inter-vehicular live video streaming for surveillance applications are presented. A test-bed for the demonstration of the achievable visual quality under different channel conditions is described.

Video transmission over IEEE 802.11p: Real-world measurements

IEEE 802.11p/ITG-G5 vehicle-to-vehicle communication technology, which enables the new class of safety and infotainment applications, is currently an emerging research topic in both industry and academia. The proposed spectrum allocation of 10 MHz channels for DSRC (Dedicated Short Range Communication) in 5.9 GHz band for the USA and Europe allows considering the transmission of video information between vehicles as one of the grounding blocks for future automotive applications. Although several published works addressed the problems of video content delivery in VANETs (Vehicular Ad-hoc NETworks), no work has been reported on real-world measurements of visual quality for video being transmitted over the IEEE 802.11p vehicle-to-vehicle communication channel. This paper presents a real-time scalable video codec as well as the first results of visual quality measurements for the video information transmitted using the off-the-shelf Componentality FlexRoad DSRC equipment.

An Efficient Adaptive Binary Range Coder and Its VLSI Architecture

In this paper, we propose a new hardware-efficient adaptive binary range coder (ABRC) and its very-large-scale integration (VLSI) architecture. To achieve this, we follow an approach that allows to reduce the bit capacity of the multiplication needed in the interval division part and shows how to avoid the need to use a loop in the renormalization part of ABRC. The probability estimation in the proposed ABRC is based on a lookup table free virtual sliding window. To obtain a higher compression performance, we propose a new adaptive window size selection algorithm. In comparison with an ABRC with a single window, the proposed system provides a faster probability adaptation at the initial encoding/decoding stage, and more accurate probability estimation for very low entropy binary sources. We show that the VLSI architecture of the proposed ABRC attains a throughput of 105.92 MSymbols/s on the FPGA platform, and consumes 18.15 mW for the dynamic part power. In comparison with the state-of-the-art MQ-coder (used in JPEG2000 standard) and the M-coder (used in H.264/Advanced Video Coding and H.265/High Efficiency Video Coding standards), the proposed ABRC architecture provides comparable throughput, reduced memory, and power consumption. Experimental results obtained for a wavelet video codec with JPEG2000-like bit-plane entropy coder show that the proposed ABRC allows to reduce the bit rate by 0.8%-8% in comparison with the MQ-coder and from 1.0%-24.2% in comparison with the M-coder.

Tensor-based linear dynamical systems for action recognition from 3D skeletons

Abstract Recent years have seen a growth in interest in skeleton-based human behavior recognition. Skeleton sequences can be expressed naturally as high-order tensor time series, and in this paper we report on the modeling and analysis of such time series using a linear dynamical system (LDS). Owing to their relative simplicity and efficiency, LDSs are the most common tool used in various disciplines for encoding spatiotemporal time series data. However, conventional LDSs process the latent and observed states at each frame of a video as a column vector, a representation that fails to take into account valuable structural information associated with human action. To correct this, we propose a tensor-based linear dynamical system (tLDS) for modeling tensor observations in time series and employ Tucker decomposition to estimate the parameters of the LDS model as action descriptors. In this manner, an action can be expressed as a subspace corresponding to a point on a Grassmann manifold on which classification can be performed using dictionary learning and sparse coding. Experiments using the MSR Action3D, UCF Kinect, and Northwestern-UCLA Multiview Action3D datasets demonstrate the excellent performance of our proposed method.

Adaptive Window Size Selection for Efficient Probability Estimation in Binary Range Coder of the 3-D DWT Video Codec

In this paper we propose a new adaptive window size selection algorithm for efficient probability estimation in binary range coder. The proposed algorithm was embedded in low-complexity video codec based on three-dimensional discrete wavelet transform (3-D DWT). Simulation results show that for the basic 3-D DWT codec the proposed algorithm provides 0.2–1.2dB quality increase for a given bit rate at the price of the computational complexity increase from 1.4 to 2 times. Herewith, the modified 3-D DWT is from 1.6 to 5 times less complex for the same quality level compared to fast software implementation of the H.264/AVC standard (x264 codec in ultrafast mode). Therefore, it is more preferable than H.264/AVC if a low-complexity scalable video coding is needed.

Poster: Target packet loss selection for inter-packet loss protection for video streaming over VANETs

This work is dedicated to a live video streaming in vehicular networks where the application-based inter-packet forward error correction is used for the packet loss protection. A practical way to apply such a protection at the video encoder side for the unknown channel loss rate is proposed. Experimental results which demonstrate the performance for the proposed approach for a scalable extension of the H.264/AVC standard (H.264/SVC) and for the non-standardized video codec based on three-dimensional discrete wavelet transform (3-D DWT) are provided.

A real-time simulcast multi-view wavelet video coding based on skipping of spatial subbands

In this paper a real-time simulcast multi-view video coding based on three-dimensional discrete wavelet transform (3D DWT) is considered. An efficient rate-distortion criterion of skipping spatial subbands is proposed. A processing of subbands is done in a group of frames from low frequency to high-frequency temporal subbands and from low frequency to high-frequency spatial subbands. If for the processed subband in the current view it appears to be more efficient not to include the highest significant bit-plane into the output bit stream, then all the corresponding temporal and spatial child subbands are skipped without any calculations of 2-D wavelet transforms and entropy encoding. Moreover, all corresponding spatial subbands in sequel views (with its child subbands) are skipped as well. Simulations results have demonstrated that the 3-D DWT codec with the proposed skipping rule has much lower computational complexity (from 2 up to 8 times) for the same quality level compared to the H.264/AVC standard in the low complexity mode.

A zero block detection algorithm used in EBCOT and its architecture

We propose a zero block detection algorithm and architecture in EBCOT. After the detailed analysis of wavelet coefficients’ precision and distribution in JPEG2000, there are three main modes of zero coefficients in the wavelet domain, i.e. zero column, zero stripe and zero code block. And we also discover that the coding information of each bit plane and the corresponding passes can be obtained simultaneously in the hardware structure. Therefore, bit plane-parallel and pass-parallel coding with zero detection is proposed, and its VLSI architecture is shown in details. The analysis and the corresponding software/hardware experimental results show that the proposed architecture reduces the processing time greatly compared with others.