KLA (United States)

With the maturity of 3D capture technology, the explosive growth of point cloud data has burdened the storage and transmission process. Traditional hybrid point cloud compression (PCC) tools relying on handcrafted priors have limited compression performance and are increasingly weak in addressing the burden induced by data growth. Recently, deep learning-based PCC methods have been introduced to continue to push the PCC performance boundary. With the thriving of deep PCC, the community urgently demands a systematic overview to conclude the past progress and present future research directions. In this paper, we have a detailed review that covers popular point cloud datasets, algorithm evolution, benchmarking analysis, and future trends. Concretely, we first introduce several widely-used PCC datasets according to their major properties. Then the algorithm evolution of existing studies on deep PCC, including lossy ones and lossless ones proposed for various point cloud types, is reviewed. Apart from academic studies, we also investigate the development of relevant international standards (i.e., MPEG standards and JPEG standards). To help have an in-depth understanding of the advance of deep PCC, we select a representative set of methods and conduct extensive experiments on multiple datasets. Comprehensive benchmarking comparisons and analysis reveal the pros and cons of previous methods. Finally, based on the profound analysis, we highlight the challenges and future trends of deep learning-based PCC, paving the way for further study.

This paper introduces novel connectivity regularization techniques for triangle-dominated meshes aimed at improving lossless compression efficiency. It addresses the irregular connectivity that arises from the triangulation of quad meshes, where vertex degrees typically range from four to eight. The proposed regularization detects triangle pairs that constitute the original quads and applies one of two regularization techniques. The first technique flips diagonal edges of certain quads to achieve consistent diagonal edge orientations. The second technique restores the original quads by removing their diagonal edges. This results in a connectivity where most vertices are surrounded either by six triangles or by four quads. To ensure lossless compression, the affected edge data can be encoded in the bitstream, allowing the decoder to restore the input mesh connectivity. The results demonstrate compression gains for meshes with position, texture coordinate, and normal attributes. Adaptively applying the two techniques yielded an average compression gain of 1.9% when preserving the affected edge data for lossless compression and 5.9% when discarding the affected edge data.

In natural language generation, language models, particularly those based on decoder-only architectures as in popular Large Language Models (LLMs), have demonstrated impressive performance across a wide range of tasks. However, encoder-decoder architectures remain highly effective for tasks involving non-text data, such as images and time-series data. The decoder relies on the attention mechanism to efficiently extract information from the encoder. While it is common practice to draw information from only the last encoder layer, this might lead to insufficient training of the encoder layer stack due to the hierarchy bypassing problem. In this work, we propose layer-wise multi-view decoding for improved encoder-decoder language models, where for each decoder layer, together with the representations from the last encoder layer, which serve as a global view, those from other encoder layers are supplemented for a stereoscopic view of the source inputs. Systematic experiments and analyses show that we successfully address the hierarchy bypassing problem, require almost negligible parameter increase, and improve the performance of sequence learning with deep representations on diverse tasks, i.e., machine translation, abstractive summarization, image captioning, video captioning, medical report generation, and paraphrase generation. In particular, our approach achieves new state-of-the-art results on benchmark datasets, including a low-resource machine translation dataset and low-resource medical report generation datasets.

Neural implicit functions have brought impressive advances to the state-of-the-art of clothed human digitization from multiple or even single images. However, despite the progress, current arts still have difficulty generalizing to unseen images with complex cloth deformation and body poses. In this work, we present GarVerseLOD, a new dataset and framework that paves the way to achieving unprecedented robustness in high-fidelity 3D garment reconstruction from a single unconstrained image. Inspired by the recent success of large generative models, we believe that one key to addressing the generalization challenge lies in the quantity and quality of 3D garment data. Towards this end, GarVerseLOD collects 6,000 high-quality cloth models with fine-grained geometry details manually created by professional artists. In addition to the scale of training data, we observe that having disentangled granularities of geometry can play an important role in boosting the generalization capability and inference accuracy of the learned model. We hence craft GarVerseLOD as a hierarchical dataset with levels of details (LOD) , spanning from detail-free stylized shape to pose-blended garment with pixel-aligned details. This allows us to make this highly under-constrained problem tractable by factorizing the inference into easier tasks, each narrowed down with smaller searching space. To ensure GarVerseLOD can generalize well to in-the-wild images, we propose a novel labeling paradigm based on conditional diffusion models to generate extensive paired images for each garment model with high photorealism. We evaluate our method on a massive amount of in-the-wild images. Experimental results demonstrate that GarVerseLOD can generate standalone garment pieces with significantly better quality than prior approaches while being robust against a large variation of pose, illumination, occlusion, and deformation. Code and dataset are available at garverselod.github.io.

Community service activities in Meunasah Teungoh Village aim to increase community participation in the Development Planning Deliberation (Musrenbang) through the use of digital technology, especially WhatsApp groups. This study adopted a participatory approach involving all elements of society, including traditional leaders, youth, and women, to ensure inclusivity in the development planning process. The methods used included an initial survey to identify the level of community awareness and digital literacy, digital literacy training, and collecting aspirations through the polling feature in the WhatsApp group. The results of the activities showed a significant increase in community participation, from 40% in 2023 to 72% in 2024, with a total of 58 proposals collected. The community satisfaction survey also showed that 85% of residents were satisfied with the transparency of information conveyed through the WhatsApp group. However, 15% of residents were dissatisfied, indicating room for improvement. This study concludes that the use of digital technology can strengthen community participation in the development planning process and recommends the need for mitigation strategies to address dissatisfaction and increase accessibility for all residents. This activity is expected to be a model for other villages in implementing the principles of participatory and sustainable development.

We present the characteristics of n-on-p frontjunction (FJ) and rear-heterojunction (RHJ) GaInP solar cells at operating temperatures (T) up to 300 °C. Photovoltaic cells with efficient operation at high T may be important for satellite missions near the sun or as laser power converters for sensors operating in harsh environments. In this article, we show that the time-resolved photoluminescence lifetime (τ TRPL ) values in both lattice-matched (LM) n-Ga 0.51 In 0.49 P and metamorphic (MM) n-Ga 0.37 In 0.63 P double heterostructures are high and increase significantly with T. In contrast, the τ TRPL values in their p-type counterparts are lower and decrease with T. We go on to demonstrate both LM and MM solar cells in FJ and RHJ configurations. The internal quantum efficiency (IQE) of MM RHJ cells increases significantly up to T = 300 °C due to the increase in both τ TRPL and the linear increase in diffusivity with T. In contrast, the IQE for MM FJ cells is nearly unchanged as T increases, while the IQE of LM cells drops sharply across all wavelengths. RHJ cells maintain higher open-circuit voltage and fill factor than their FJ counterparts, leading to a significant efficiency advantage at T = 100-300 °C. Taken together, our work shows that MM cells perform well at elevated T and that RHJ cells are promising for high-T operation.

Introdução: A criança com acometimento neurológico pode apresentar comprometimentos nas habilidades motoras, alterações sensoriais, déficits de comunicação, linguagem e alterações no padrão miccional e evacuatório. Objetivo: Relatar os efeitos da intervenção transdisciplinar na participação ativa da criança com deficiência, no manejo do intestino. Metodologia: Trata-se de um relato de caso de uma criança, que participou de um protocolo de reabilitação transdisciplinar no manejo do intestino, aprovado pelo comitê de ética (CAAE-74512023.8.0000.0129), realizado no Núcleo de Reabilitação Infantil, NeuroAmar, Natal/RN. Resultados: Criança, 4 anos e 5 meses de idade, diagnóstico clínico de Fenda labial com fenda palatina (CID, S: CID-10-037), Outras malformações congênitas do cérebro (CID-10-204), Retardo do desenvolvimento fisiológico normal (CID-10-R62), Epilepsia e síndromes epilépticas sintomáticas definidas por sua localização (focal) e (parcial) com crises parciais complexas (CID-10-G40.2) e Paralisia cerebral (CID-10 G 80). A queixa principal da mãe da criança foi a ausência de comunicação do desejo de evacuar e poder ser levado ao vaso sanitário. A criança não apresenta alteração na função intestinal, no entanto, se alimenta via sonda de gastrostomia, se comunica de forma alternativa por meio do sistema de comunicação alternativa e aumentativa (PODD12), como também, apresenta acometimento motor global que interfere diretamente em suas habilidade motoras e funcionais. A intervenção transdisciplinar foi realizada por quatro horas seguidas diariamente no período de 3 meses, a fisioterapia teve como objetivo, melhorar a permanência em sedestação com bom controle de tronco, no setting da Terapia Ocupacional foi trabalhado a função manual e as percepções sensoriais, no setting da Fonoaudiologia, a comunicação do desejo evacuatório. Após a intervenção, a criança passou a relatar suas percepções sobre a evacuação por meio do PODD12, e a levantar os braços para sinalizar o desejo de evacuar, dessa forma podendo ser levado ao vaso sanitário. Conclusão: No caso relatado, fica evidente o efeito positivo da intervenção transdisciplinar para a participação ativa da criança com deficiência no manejo do intestino, e o ganho de capacidade e participação. No entanto, foi visto a falta de um instrumento específico de avaliação para mensurar a evolução diária e longitudinal da criança.

[SPA] El desarrollo de nuevos enfoques que aseguren la calidad de las hortalizas depende de una mejor comprensión de su fisiología y bioquímica, así como de la influencia del ambiente y la tecnología de producción. El objetivo de este estudio fue determinar la incidencia de la dosis de fertilizante en un cultivo de lechuga a campo sobre la calidad del producto de acuerdo a la época de producción y el material genético empleado. Los ensayos se realizaron en otoño-invierno y primavera con lechuga mantecosa (cv. Daguan) y de hojas sueltas (cv. Brisa) empleando un manejo de cultivo convencional. Los tratamientos fueron: lombricompuesto (24 t ha-1), 75 y 150 kg N ha-1 aplicado como urea. A cosecha fueron seleccionadas las hojas externas e internas, lavadas en soluciones clorinadas, pesadas, envasadas en bolsas de poliolefina y almacenadas a 1ºC durante siete días. Los muestreos se tomaron a cosecha y cada dos días durante el almacenamiento para medir las concentraciónes de nitratos y ácido ascórbico. La fertilización nitrogenada incrementó la acumulación de nitratos, especialmente en el cultivo otoño-invernal independientemente del genotipo. En general, las hojas externas presentaron una menor concentración de nitratos que las hojas internas. Los nitratos disminuyeron al final del almacenamiento. Los niveles de ácido ascórbico a cosecha fueron en general mayores en otoño para Brisa, en cambio en primavera Daguan tuvo mayores niveles en hojas externas con fertilización. El aumento de los niveles de fertilización en primavera tendió a disminuir el ácido ascórbico a cosecha en Brisa. En el otoño este cultivar tuvo más ácido ascórbico con 150 kg N ha-1. Por otro lado, Daguan tendió a mayores niveles de ácido ascórbico con mayor nivel de fertilización en ambas épocas, particularmente en hojas internas. Al final del almacenamiento el ácido ascórbico en Brisa no tuvo una tendencia definida, mientras que en Daguan aumentó. [ENG] The development of new approaches to assure the quality of vegetables depends on a better understanding of the physiology and biochemistry, and also the influence of the environment and the production technology. The objective of this study was to determine the incidence of the fertilizer dose on a field grown lettuce crop over the quality of the product in relation to the season production and the genetic material used. The experiments were carried on fall-winter and spring with butterhead lettuce (cv. Daguan) and leafy lettuce (cv. Brisa) using a conventional system of production. Treatment were vermicompost (24 t ha-1), 75 y 150 kg N ha-1 applied as urea. At harvest internal and external leaves were selected, washed in chlorine solutions, weighted, packed in polyolefin bags and stored at 1ºC during seven days. Samples were taken at harvest and each two days during the storage to measure nitrate and ascorbic acid concentrations. The N fertilization increased nitrate accumulation, especially in the fall-winter crop independently of the genotype. In general, the external leaves showed lower nitrate concentrations than internal leaves. Nitrate decreased at the end of the storage. Ascorbic acid levels at harvest were in general higher for Brisa during fall-winter, but in spring Daguan had higher levels in external leaves with fertilization. Higher levels of N fertilization in spring made ascorbic acid concentrations in Brisa tend to decrease. This cultivar in fall-winter had more ascorbic acid with 150 Kg N ha-1. On the other hand, Daguan tended to higher levels of ascorbic acid with higher levels of fertilization in both seasons of growth, especially in internal leaves. At the end of the storage ascorbic acid of Brisa had no a defined trend while Daguan increased.