Abstract: Autonomous vehicles offer the possibility of significant benefits to social welfare. However, fully automated cars might not be going to happen in the near further. To speed the adoption of the self-driving technologies, many governments worldwide are passing laws requiring data recorders for the testing of autonomous vehicles. Currently, the self-driving vehicle, (e.g., shuttle bus) has to be monitored from a remote control center. When an autonomous vehicle encounters an unexpected driving environment, such as road construction or an obstruction, it should request assistance from a remote operator. Nevertheless, large amounts of data, including images, radar and lidar data, etc., have to be transmitted from the vehicle to the remote center. Therefore, this paper proposes a data compression method of in-vehicle networks for remote monitoring of autonomous vehicles. Firstly, the time-series data are rearranged into a multi-dimensional signal space. Upon the arrival, for controller area networks (CAN), the new data are mapped onto a time-data two-dimensional space associated with the specific CAN identity. Secondly, the data are sampled based on differential sampling. Finally, the whole set of data are encoded using existing algorithms such as Huffman, arithmetic and codebook encoding methods. To evaluate system performance, the proposed method was deployed on an in-house built autonomous vehicle. The testing results show that the amount of data can be reduced as much as 1/7 compared to the raw data.
Abstract: Background modeling and subtraction in video
analysis has been widely used as an effective method for moving
objects detection in many computer vision applications. Recently, a
large number of approaches have been developed to tackle different
types of challenges in this field. However, the dynamic background
and illumination variations are the most frequently occurred problems
in the practical situation. This paper presents a favorable two-layer
model based on codebook algorithm incorporated with local binary
pattern (LBP) texture measure, targeted for handling dynamic
background and illumination variation problems. More specifically,
the first layer is designed by block-based codebook combining with
LBP histogram and mean value of each RGB color channel. Because
of the invariance of the LBP features with respect to monotonic
gray-scale changes, this layer can produce block wise detection results
with considerable tolerance of illumination variations. The pixel-based
codebook is employed to reinforce the precision from the output of the
first layer which is to eliminate false positives further. As a result, the
proposed approach can greatly promote the accuracy under the
circumstances of dynamic background and illumination changes.
Experimental results on several popular background subtraction
datasets demonstrate very competitive performance compared to
previous models.
Abstract: Digital cameras to reduce cost, use an image sensor to
capture color images. Color Filter Array (CFA) in digital cameras
permits only one of the three primary (red-green-blue) colors to be
sensed in a pixel and interpolates the two missing components
through a method named demosaicking. Captured data is interpolated
into a full color image and compressed in applications. Color
interpolation before compression leads to data redundancy. This
paper proposes a new Vector Quantization (VQ) technique to
construct a VQ codebook with Differential Evolution (DE)
Algorithm. The new technique is compared to conventional Linde-
Buzo-Gray (LBG) method.
Abstract: Because current wireless communication requires high
reliability in a limited bandwidth environment, this paper proposes
the variable modulation scheme based on the codebook. The variable
modulation scheme adjusts transmission power using the codebook in
accordance with channel state. Also, if the codebook is composed of
many bits, the reliability is more improved by the proposed scheme.
The simulation results show that the performance of proposed scheme
has better reliability than the the performance of conventional scheme.
Abstract: A face recognition system is a computer application for automatically identifying or verifying a person from a digital image or a video frame. A lot of algorithms have been proposed for face recognition. Vector Quantization (VQ) based face recognition is a novel approach for face recognition. Here a new codebook generation for VQ based face recognition using Integrated Adaptive Fuzzy Clustering (IAFC) is proposed. IAFC is a fuzzy neural network which incorporates a fuzzy learning rule into a competitive neural network. The performance of proposed algorithm is demonstrated by using publicly available AT&T database, Yale database, Indian Face database and a small face database, DCSKU database created in our lab. In all the databases the proposed approach got a higher recognition rate than most of the existing methods. In terms of Equal Error Rate (ERR) also the proposed codebook is better than the existing methods.
Abstract: In this paper, a joint source-channel coding (JSCC) scheme for time-varying channels is presented. The proposed scheme uses hierarchical framework for both source encoder and transmission via QAM modulation. Hierarchical joint source channel codes with hierarchical QAM constellations are designed to track the channel variations which yields to a higher throughput by adapting certain parameters of the receiver to the channel variation. We consider the problem of still image transmission over time-varying channels with channel state information (CSI) available at 1) receiver only and 2) both transmitter and receiver being informed about the state of the channel. We describe an algorithm that optimizes hierarchical source codebooks by minimizing the distortion due to source quantizer and channel impairments. Simulation results, based on image representation, show that, the proposed hierarchical system outperforms the conventional schemes based on a single-modulator and channel optimized source coding.
Abstract: Wireless Sensor Network (WSN) comprises of sensor
nodes which are designed to sense the environment, transmit sensed
data back to the base station via multi-hop routing to reconstruct
physical phenomena. Since physical phenomena exists significant
overlaps between temporal redundancy and spatial redundancy, it is
necessary to use Redundancy Suppression Algorithms (RSA) for sensor
node to lower energy consumption by reducing the transmission
of redundancy. A conventional algorithm of RSAs is threshold-based
RSA, which sets threshold to suppress redundant data. Although
many temporal and spatial RSAs are proposed, temporal-spatial RSA
are seldom to be proposed because it is difficult to determine when
to utilize temporal or spatial RSAs. In this paper, we proposed a
novel temporal-spatial redundancy suppression algorithm, Codebookbase
Redundancy Suppression Mechanism (CRSM). CRSM adopts
vector quantization to generate a codebook, which is easily used to
implement temporal-spatial RSA. CRSM not only achieves power
saving and reliability for WSN, but also provides the predictability
of network lifetime. Simulation result shows that the network lifetime
of CRSM outperforms at least 23% of that of other RSAs.
Abstract: This paper explores the implementation of adaptive
coding and modulation schemes for Multiple-Input Multiple-Output
Orthogonal Frequency Division Multiplexing (MIMO-OFDM) feedback
systems. Adaptive coding and modulation enables robust and
spectrally-efficient transmission over time-varying channels. The basic
premise is to estimate the channel at the receiver and feed this estimate
back to the transmitter, so that the transmission scheme can be
adapted relative to the channel characteristics. Two types of codebook
based channel feedback techniques are used in this work. The longterm
and short-term CSI at the transmitter is used for efficient channel
utilization. OFDM is a powerful technique employed in communication
systems suffering from frequency selectivity. Combined with
multiple antennas at the transmitter and receiver, OFDM proves to be
robust against delay spread. Moreover, it leads to significant data rates
with improved bit error performance over links having only a single
antenna at both the transmitter and receiver. The coded modulation
increases the effective transmit power relative to uncoded variablerate
variable-power MQAM performance for MIMO-OFDM feedback
system. Hence proposed arrangement becomes an attractive approach
to achieve enhanced spectral efficiency and improved error rate
performance for next generation high speed wireless communication
systems.
Abstract: This paper investigates the performance of Multiple- Input Multiple-Output (MIMO) feedback system combined with Orthogonal Frequency Division Multiplexing (OFDM). Two types of codebook based channel feedback techniques are used in this work. The first feedback technique uses a combination of both the long-term and short-term channel state information (CSI) at the transmitter, whereas the second technique uses only the short term CSI. The long-term and short-term CSI at the transmitter is used for efficient channel utilization. OFDM is a powerful technique employed in communication systems suffering from frequency selectivity. Combined with multiple antennas at the transmitter and receiver, OFDM proves to be robust against delay spread. Moreover, it leads to significant data rates with improved bit error performance over links having only a single antenna at both the transmitter and receiver. The effectiveness of these techniques has been demonstrated through the simulation of a MIMO-OFDM feedback system. The results have been evaluated for 4x4 MIMO channels. Simulation results indicate the benefits of the MIMO-OFDM channel feedback system over the one without incorporating OFDM. Performance gain of about 3 dB is observed for MIMO-OFDM feedback system as compared to the one without employing OFDM. Hence MIMO-OFDM becomes an attractive approach for future high speed wireless communication systems.
Abstract: In this paper we propose segmentation approach based
on Vector Quantization technique. Here we have used Kekre-s fast
codebook generation algorithm for segmenting low-altitude aerial
image. This is used as a preprocessing step to form segmented
homogeneous regions. Further to merge adjacent regions color
similarity and volume difference criteria is used. Experiments
performed with real aerial images of varied nature demonstrate that
this approach does not result in over segmentation or under
segmentation. The vector quantization seems to give far better results
as compared to conventional on-the-fly watershed algorithm.
Abstract: A low bit rate still image compression scheme by
compressing the indices of Vector Quantization (VQ) and generating
residual codebook is proposed. The indices of VQ are compressed by
exploiting correlation among image blocks, which reduces the bit per
index. A residual codebook similar to VQ codebook is generated that
represents the distortion produced in VQ. Using this residual
codebook the distortion in the reconstructed image is removed,
thereby increasing the image quality. Our scheme combines these two
methods. Experimental results on standard image Lena show that our
scheme can give a reconstructed image with a PSNR value of 31.6 db
at 0.396 bits per pixel. Our scheme is also faster than the existing VQ
variants.
Abstract: This paper presents a very simple and efficient
algorithm for codebook search, which reduces a great deal of
computation as compared to the full codebook search. The algorithm
is based on sorting and centroid technique for search. The results
table shows the effectiveness of the proposed algorithm in terms of
computational complexity. In this paper we also introduce a new
performance parameter named as Average fractional change in pixel
value as we feel that it gives better understanding of the closeness of
the image since it is related to the perception. This new performance
parameter takes into consideration the average fractional change in
each pixel value.
Abstract: In this paper, a new algorithm for generating codebook is proposed for vector quantization (VQ) in image coding. The significant features of the training image vectors are extracted by using the proposed Orthogonal Polynomials based transformation. We propose to generate the codebook by partitioning these feature vectors into a binary tree. Each feature vector at a non-terminal node of the binary tree is directed to one of the two descendants by comparing a single feature associated with that node to a threshold. The binary tree codebook is used for encoding and decoding the feature vectors. In the decoding process the feature vectors are subjected to inverse transformation with the help of basis functions of the proposed Orthogonal Polynomials based transformation to get back the approximated input image training vectors. The results of the proposed coding are compared with the VQ using Discrete Cosine Transform (DCT) and Pairwise Nearest Neighbor (PNN) algorithm. The new algorithm results in a considerable reduction in computation time and provides better reconstructed picture quality.
Abstract: The speech signal conveys information about the
identity of the speaker. The area of speaker identification is
concerned with extracting the identity of the person speaking the
utterance. As speech interaction with computers becomes more
pervasive in activities such as the telephone, financial transactions
and information retrieval from speech databases, the utility of
automatically identifying a speaker is based solely on vocal
characteristic. This paper emphasizes on text dependent speaker
identification, which deals with detecting a particular speaker from a
known population. The system prompts the user to provide speech
utterance. System identifies the user by comparing the codebook of
speech utterance with those of the stored in the database and lists,
which contain the most likely speakers, could have given that speech
utterance. The speech signal is recorded for N speakers further the
features are extracted. Feature extraction is done by means of LPC
coefficients, calculating AMDF, and DFT. The neural network is
trained by applying these features as input parameters. The features
are stored in templates for further comparison. The features for the
speaker who has to be identified are extracted and compared with the
stored templates using Back Propogation Algorithm. Here, the
trained network corresponds to the output; the input is the extracted
features of the speaker to be identified. The network does the weight
adjustment and the best match is found to identify the speaker. The
number of epochs required to get the target decides the network
performance.
Abstract: In this paper, we propose an improved 3D star skeleton
technique, which is a suitable skeletonization for human posture representation
and reflects the 3D information of human posture.
Moreover, the proposed technique is simple and then can be performed
in real-time. The existing skeleton construction techniques, such as
distance transformation, Voronoi diagram, and thinning, focus on the
precision of skeleton information. Therefore, those techniques are not
applicable to real-time posture recognition since they are computationally
expensive and highly susceptible to noise of boundary. Although
a 2D star skeleton was proposed to complement these problems,
it also has some limitations to describe the 3D information of the
posture. To represent human posture effectively, the constructed skeleton
should consider the 3D information of posture. The proposed 3D
star skeleton contains 3D data of human, and focuses on human action
and posture recognition. Our 3D star skeleton uses the 8 projection
maps which have 2D silhouette information and depth data of human
surface. And the extremal points can be extracted as the features of 3D
star skeleton, without searching whole boundary of object. Therefore,
on execution time, our 3D star skeleton is faster than the “greedy" 3D
star skeleton using the whole boundary points on the surface. Moreover,
our method can offer more accurate skeleton of posture than the
existing star skeleton since the 3D data for the object is concerned.
Additionally, we make a codebook, a collection of representative 3D
star skeletons about 7 postures, to recognize what posture of constructed
skeleton is.
Abstract: The paper presents a complete discrete statistical framework, based on a novel vector quantization (VQ) front-end process. This new VQ approach performs an optimal distribution of VQ codebook components on HMM states. This technique that we named the distributed vector quantization (DVQ) of hidden Markov models, succeeds in unifying acoustic micro-structure and phonetic macro-structure, when the estimation of HMM parameters is performed. The DVQ technique is implemented through two variants. The first variant uses the K-means algorithm (K-means- DVQ) to optimize the VQ, while the second variant exploits the benefits of the classification behavior of neural networks (NN-DVQ) for the same purpose. The proposed variants are compared with the HMM-based baseline system by experiments of specific Arabic consonants recognition. The results show that the distributed vector quantization technique increase the performance of the discrete HMM system.
Abstract: In this paper, we propose a novel improvement for the generalized Lloyd Algorithm (GLA). Our algorithm makes use of an M-tree index built on the codebook which makes it possible to reduce the number of distance computations when the nearest code words are searched. Our method does not impose the use of any specific distance function, but works with any metric distance, making it more general than many other fast GLA variants. Finally, we present the positive results of our performance experiments.