The operation of a complete license plate recognition system includes three parts: license plate localization, character segmentation, and character identification. Among these three parts, license plate localization is relatively more difficult and complicated. Until now, differentiating background and real license plate images in real and random traffic conditions remains to be a very difficult task. Via a VQ coding technique, this study introduces a method resolve this problem. As a preprocessing step, this method first converts an image to be classified into binary form by using statistics generated from a license plate image database. The next step of the proposed approach is to use a VQ method to represent the image by a series of codewords. By computing the probability of these codewords used by the license plate and background images, these codewords are renumbered. By using neural networks to classify such images, our experimental results show that the proposed approach can differentiate background and real license plate images with a very high successful rate.

Abstract With the rapid progress of the MEMs process, the cost and the size of accelerometers are reducing rapidly. As a result, accelerometers have found many new applications in industrial, entertainment and medical domains. One of such an applications is to acquire information about human body movement. The objective of this work is to use knee acceleration signal for indentity verification. Comparing with traditional biometric methods, this approach has several distinct features. First, it can aquire a large amount of data efficiently and conventiently. Second, it is relatively difficult to duplicate. In designing the verification algorithm, this study has developed a neural network method a hyperspherical classifier method. The experimental results demonstrated that hyperspherical classifier provide better performances in this application. By setting the sensitively to 85%, the specificity achieved by the hyperspherical classifier is at least 95%.

This thesis proposes a distributed beamforming technique in wireless networks with half-duplex amplify-and-forward relays. With full channel state information, it has been shown that transmit beamforming is able to achieve significant diversity and coding gain. However, it takes large amount of overhead. First, we adopt the Generalized Lloyd Algorithm to design codebooks which minimize average SNR, and reduce the feedback rate by quantizing the channel state information. Furthermore, we utilize the correlation property of time-varying channels to compress the size of feedback message required to accomplish distributed beamforming. We model time-varying channels as a first-order finite-state Markov chain, namely the emph{channel state Markov chain}. Then, we propose two methods to compress the feedback bits according to the property of the transition probabilities among different channel states. One method is to compress the feedback by discarding some channel states which is less likely to be transited given current state. In the other method, we reserve all channel states and adopt Huffman coding to compress the feedback bits based on the transition probabilities. Simulations also show that distributed beamforming with compressed feedback performs closely to the beamforming with infinite feedback.

Designing the Nearest Neighbor Classifiers via the VQ Method

One primary topic of sleep studies is the depth of sleep. According to definitions of R&K rules, human sleep can be roughly divided into three different stages: Awake, Non-rapid-eye-movement (NREM) Sleep, and Rapid-eye-movement (REM) Sleep. Moreover, sleep stages are scored mainly by EEG signals and complementally by EOG and EMG signals. Many researchers have indicated that diseases or disorders occur during sleep will affect life quality of patients. For example, REM sleep-related dyssomnia is highly correlated with neurodegenerative or mental disorders such as major depression. Furthermore, sleep apnea is one of the most common sleep disorders at present. Untreated sleep apnea can increase the risk of mental and cardiovascular diseases. This research proposes a detection method of REM sleep. Take into account the environment of homecare, we just extract and analyze EOG signals for the sake of convenience in comparison with EEG channels. By analyzing elementary waveforms of EOG signals based on VQ method, the proposed method performs a classification accuracy of 67.71% in a group application. The corresponding sensitivity and specificity are 73.38% and 68.95% respectively. In contrast, the average classification accuracy is 82.02% in personalized applications. And the corresponding average sensitivity and specificity are 83.05% and 81.62% respectively. Experimental results demonstrate the feasibility of detecting REM sleep via the proposed method, especially in personalized applications. This will be propitious to a long term tracing and research of personal sleep status.

The rapid-eye-movement (REM) sleep is one of the most important parts in overnight sleep. In this study, an automatic REM sleep staging rule is introduced. Compared with the traditional REM detection method, a distinct feature of this method is that it only requires two EOG signals and thus reduces the number of input signal channels significantly. We calculate the correlation coefficient series between two EOG signals. By representing such a series with a VQ coding method, several techniques are proposed to improve the classification rate. Experimental results are given to demonstrate the effectiveness of the proposed approach.

[ES] Las especies reactivas de oxígeno (ROS), el oxígeno molecular (O2) y el óxido nítrico (NO) son factores comunes en las respuestas a diferentes tipos de estrés en las plantas. En los últimos años, las condiciones ambientales adversas han aumentado debido al cambio climático y, entre ellas, las fuertes lluvias que provocan anegaciones o inundaciones en los campos suponen un grave problema para el desarrollo y el futuro de la agricultura. Estas inundaciones provocan que la planta se enfrente a condiciones de hipoxia severas y a la re-oxigenación que ocurre cuando se retira el agua. Es importante, por tanto, conocer los factores genéticos y moleculares que modulan las respuestas de las plantas a la hipoxia u otros tipos de estrés abiótico para hacer frente a estas situaciones. La familia de proteínas VQ de Arabidopsis thaliana incluye cinco miembros codificados por genes que se inducen en respuesta a la hipoxia, al NO y al estrés oxidativo. Dos de ellos, VQ1 y VQ10, codifican dos proteínas con una gran homología de secuencia y estructura, que no poseen actividad transcripcional. En este trabajo, se describe que VQ1 y VQ10 interaccionan entre sí y con ellas mismas, y con las otras tres proteínas VQ y cuatro factores de transcripción WRKY cuyos genes también se inducen en respuesta a hipoxia, NO y estrés oxidativo. Un rastreo de doble híbrido de levadura de una genoteca de Arabidopsis utilizando VQ1 como cebo permitió identificar alrededor de 300 proteínas interactoras de VQ1 y probablemente también de VQ10, sugiriendo que ambas proteínas tienen patrones similares de interacción proteína-proteína. VQ1 y VQ10 son proteínas citoplásmicas y nucleares, por lo que la identificación de casi 50 proteínas cloroplásticas que interaccionan con VQ1 en el cloroplasto sugiere que estas interacciones deben ocurrir en el citoplasma. En el rastreo se identificó la 1-desoxi-D-xilulosa-5-fosfato sintasa (DXS), una enzima clave para la biosíntesis de isoprenoides, la fotosíntesis y el crecimiento de las plantas. Al co-inmunoprecipitar DXS y VQ1 o VQ10 se observó que VQ1 y sobre todo VQ10 favorecieron la monomerización de DXS en un ambiente reductor. A partir de la identificación de un mutante hipermórfico vq10-H se ha procedido a caracterizar algunas de las funciones reguladoras ejercidas por VQ10. El análisis transcriptómico comparado de vq10-H y plantas silvestres Col-0 permitió identificar la regulación ejercida por VQ10 en el desarrollo de las raíces, en las respuestas a NO y estrés oxidativo, así como una potencial función reguladora del procesamiento alternativo de ARNs. La sobreexpresión de VQ10 provocó el aumento de la elongación de la raíz primaria, y de la división y elongación celular en la raíz. También provocó una mayor tolerancia al estrés oxidativo, una menor sensibilidad al NO, y una mayor eficiencia fotosintética en comparación con las plantas silvestres. En conjunto, el trabajo realizado en esta Tesis sugiere que algunas funciones reguladoras de VQ1 y VQ10 en el desarrollo y las respuestas a estrés de las plantas podría estar basado en mecanismos de regulación redox. / [CA] Les espècies reactives d'oxigen (ROS), l'oxigen molecular (O2) i l'òxid nítric (NO) són factors comuns en les respostes a diferents tipus d'estrés en les plantes. En els últims anys, les condicions ambientals adverses han augmentat a causa del canvi climàtic i, entre elles, les fortes pluges que provoquen inundacions als camps suposen un greu problema per al desenvolupament i el futur de l'agricultura. Aquestes inundacions provoquen que la planta s'enfronte a condicions d'hipòxia severes i a la re-oxigenació que ocorre quan es retira l'aigua. És important, per tant, conéixer els factors genètics i moleculars que modulen les respostes de les plantes a l'hipòxia o altres tipus d'estrés abiòtic per a fer front a aquestes situacions. La família de proteïnes VQ d'Arabidopsis thaliana inclou cinc membres codificats per gens que s'indueixen com a resposta a l'hipòxia, al NO i a l'estrés oxidatiu. Dos d'ells, VQ1 i VQ10, codifiquen dues proteïnes amb una gran homologia de seqüència i estructura, que no posseeixen activitat transcripcional. En aquest treball, es descriu que VQ1 i VQ10 interaccionen entre si i amb elles mateixes, i amb les altres tres proteïnes VQ i quatre factors de transcripció WRKY, els gens dels quals també s'indueixen com a resposta a hipòxia, NO i estrés oxidatiu. Un rastreig amb la técnica del doble híbrid en llevat d'una genoteca d'Arabidopsis utilitzant VQ1 com a esquer va permetre identificar al voltant de 300 proteïnes interactores de VQ1 i probablement també de VQ10, suggerint que ambdues proteïnes tenen patrons similars d'interacció proteïna-proteïna. VQ1 i VQ10 són proteïnes citoplasmàtiques i nuclears, de manera que la identificació de gairebé 50 proteïnes cloroplàstiques que interaccionen amb VQ1 en el cloroplast suggereix que aquestes interaccions han d'ocorrer al citoplasma. En el rastreig es va identificar la 1-desoxi-D-xilulosa-5-fosfat sintasa (DXS), un enzim clau per a la biosíntesi d'isoprenoides, la fotosíntesi i el creixement de les plantes. En co-immunoprecipitar DXS i VQ1 o VQ10 es va observar que VQ1 i sobretot VQ10 van afavorir la monomerització de DXS en un ambient reductor. A partir de la identificación d'un mutant hipermòrfic vq10-H s'ha procedit a caracteritzar algunes de les funcions reguladores exercides per VQ10. L'anàlisi transcriptòmic comparat de vq10-H i plantes silvestres Col-0 va permetre identificar la regulació exercida per VQ10 en el desenvolupament de les arrels, en les respostes a NO i estrès oxidatiu, així com una potencial funció reguladora del processament alternatiu d'ARNs. La sobreexpressió de VQ10 va provocar l'augment de l'elongació de l'arrel primària, i de la divisió i elongació cel·lular a l'arrel. També va provocar una major tolerància a l'estrès oxidatiu, una menor sensibilitat al NO, i una major eficiencia fotosintètica en comparació amb les plantes silvestres. En conjunt, el treball realitzat en aquesta Tesi suggereix que algunes funcions reguladores de VQ1 i VQ10 en el desenvolupament i les respostes a estrés de les plantes podria estar basat en mecanismes de regulació redox. / [EN] Reactive oxygen species (ROS), molecular oxygen (O2), and nitric oxide (NO) are common factors in diverse plant stress responses. In the past few years, adverse environmental conditions have increased due to climate change, including heavy rains leading to waterlogging or flooding in fields, posing a serious problem for agriculture's development and future. Floods impose severe hypoxia conditions to plants to be later re-oxygenated when water recedes. Therefore, it is essential to understand the genetic and molecular factors that modulate plant responses to hypoxia and other abiotic stresses to address these situations. The VQ-motif containing protein family in Arabidopsis thaliana includes five members encoded by genes induced in response to hypoxia, NO, and oxidative stress. Two of them, VQ1 and VQ10, encode two proteins with significant sequence and structural homology, lacking transcriptional activity. In this study, it is described that VQ1 and VQ10 interact with each other and themselves, as well as with the other three VQ proteins and four WRKY transcription factors whose genes are also induced in response to hypoxia, NO, and oxidative stress. A yeast two-hybrid screening of an Arabidopsis library using VQ1 as bait allowed us identifying almost 300 VQ1-interacting proteins, that likely also interact with VQ10, suggesting that their homology can be extended to protein-protein interaction patterns. VQ1 and VQ10 are cytoplasmic and nuclear proteins, so the identification of nearly 50 chloroplast proteins interacting with VQ1 in the chloroplast suggests that these interactions may be occurring in the cytoplasm. 1-Deoxy-D-xylulose-5-phosphate synthase (DXS), a key enzyme in isoprenoid biosynthesis, photosynthesis, and plant growth, was identified in the screening. Co-immunoprecipitation of DXS with VQ1 or VQ10 showed that VQ1 and, particularly, VQ10 promoted the monomerization of DXS under reducing conditions. Based on the identification of a hypermorphic vq10-H mutant, some of the regulatory functions exerted by VQ10 have been characterized. Comparative transcriptomic analysis of the hypermorphic vq10-H mutant and Col-0 wild type plants allowed us identifying the regulation exerted by VQ10 in root development, in responses to NO and oxidative stress, as well as in the alternative splicing of RNAs. Overexpression of VQ10 led to increased primary root elongation, and increased root cell division and elongation. It also triggered an enhanced oxidative stress tolerance, reduced sensitivity to NO, and enhanced photosynthetic efficiency compared to wild type plants. Altogether, the work carried out in this thesis suggest that some of the regulatory functions exerted by VQ1 and VQ10 on development and stress responses in plants, could be based on redox mechanisms. / Gayubas Balaguer, B. (2023). Función de las proteínas VQ1 y VQ10 codificadas por genes inducibles por hipoxia, óxido nítrico, y estrés oxidativo en la regulación del desarrollo y las respuestas a estrés en Arabidopsis thaliana [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/202571

This thesis focuses on addressing the limitations of existing human motion prediction models by extending the prediction horizon to very long-term forecasts. The objective is to develop a model that achieves one of the best stable prediction horizons in the field, providing accurate predictions without significant error increase over time. Through the utilization of quantization based models our research successfully achieves the desired objective with the proposed aligned version of Mean Per Joint Position Error. The first of the two proposed models, an attention-based Vector Quantized Variational AutoEncoder, demonstrates good performance in predicting beyond conventional time boundaries, maintaining low error rates as the prediction horizon extends. While slight discrepancies in joint positions are observed, the model effectively captures the underlying patterns and dynamics of human motion, which remains highly applicable in real-world scenarios. Furthermore, our investigation into a correlation-based Vector Quantized Variational AutoEncoder, as an alternative to attention-based one, highlights the challenges in capturing complex relationships and meaningful patterns within the data. The correlation-based VQ-VAE’s tendency to predict flat outputs emphasizes the need for further exploration and innovative approaches to improve its performance. Overall, this thesis contributes to the field of human motion prediction by extending the prediction horizon and providing insights into model performance and limitations. The developed model introduces a novel option to consider when contemplating long-term prediction applications across various domains and sets the foundation for future research to enhance performance in long-term scenarios. / Denna avhandling fokuserar på att hantera begränsningarna i befintliga modeller för förutsägelse av mänskliga rörelser genom att utöka förutsägelsehorisonten till mycket långsiktiga prognoser. Målet är att utveckla en modell som uppnår en av de bästa stabila prognoshorisonterna inom området, vilket ger korrekta prognoser utan betydande felökning över tiden. Genom att använda kvantiseringsbaserade modeller uppnår vår forskning framgångsrikt det önskade målet med den föreslagna anpassade versionen av Mean Per Joint Position Error. Den första av de två föreslagna modellerna, en uppmärksamhetsbaserad Vector Quantized Variational AutoEncoder, visar goda resultat när det gäller att förutsäga bortom konventionella tidsgränser och bibehåller låga felfrekvenser när förutsägelsehorisonten förlängs. Även om små avvikelser i ledpositioner observeras, fångar modellen effektivt de underliggande mönstren och dynamiken i mänsklig rörelse, vilket förblir mycket tillämpligt i verkliga scenarier. Vår undersökning av en korrelationsbaserad Vector Quantized Variational AutoEncoder, som ett alternativ till en uppmärksamhetsbaserad sådan, belyser dessutom utmaningarna med att fånga komplexa relationer och meningsfulla mönster i data. Den korrelationsbaserade VQ-VAE:s tendens att förutsäga platta utdata understryker behovet av ytterligare utforskning och innovativa metoder för att förbättra dess prestanda. Sammantaget bidrar denna avhandling till området för förutsägelse av mänskliga rörelser genom att utöka förutsägelsehorisonten och ge insikter om modellens prestanda och begränsningar. Den utvecklade modellen introducerar ett nytt alternativ att ta hänsyn till när man överväger långsiktiga prediktionstillämpningar inom olika områden och lägger grunden för framtida forskning för att förbättra prestanda i långsiktiga scenarier.

<p>This thesis explores the possibilities of avoiding the issues generally associated with compression of noisy imagery, through the usage of vector quantization. By utilizing the learning aspects of vector quantization, image processing operations such as noise reduction could be implemented in a straightforward way. Several techniques are presented and evaluated. A direct comparison shows that for noisy imagery, vector quantization, in spite of it's simplicity, has clear advantages over MPEG-4 encoding.</p>

This thesis explores the possibilities of avoiding the issues generally associated with compression of noisy imagery, through the usage of vector quantization. By utilizing the learning aspects of vector quantization, image processing operations such as noise reduction could be implemented in a straightforward way. Several techniques are presented and evaluated. A direct comparison shows that for noisy imagery, vector quantization, in spite of it's simplicity, has clear advantages over MPEG-4 encoding.