Search

Accurate and efficient estimation of animal poses is essential in various fields, including animal behavior research, pharmaceutical studies, and bionic robotics. This capability provides immediate feedback and enables rapid decision-making in practical applications. However, pose estimation typically relies on numerous learnable parameters and complex model architectures to achieve high accuracy, making real-time detection challenging, especially in scenarios with low computational power. This paper proposes an efficient animal pose estimation network (EAPoseNet) for resource-constrained environments that achieve real-time detection without compromising accuracy. EAPoseNet designs a more advanced feature extraction network (EAPBackbone) and a key point localization network (EAPHead) for animal pose estimation tasks. We conducted multiple comparative experiments on the AP-10K, Desert Locust, Vinegar Fly, Zebra, Mouse, and Horse10 datasets to evaluate EAPoseNet. Additionally, we performed ablation studies on the Mouse dataset. The experimental results demonstrate that EAPoseNet delivers competitive performance with a minimal number of parameters. On the AP-10K dataset, our method’s average precision (AP) is 1.3% higher than that of the RTMPose-M model of the same scale, while EAPoseNet has 1.3 million fewer parameters than RTMPose-M. On the Mouse dataset, the best implementation of EAPoseNet enhances the model’s AP by 0.1% to 0.5%. [ABSTRACT FROM AUTHOR]

To face with the surge of cloud computing services, energy problem and carbon mission constraint, more and more geo-distributed hybrid powered data centers are build over the world. Geo-distributed hybrid powered data centers are located in different locations equipped with different types of green energy source. In this work, we propose a periodic requests dispatcher, including a periodic model and a spatial balance algorithm, to minimize the brown energy use. The periodic model divides the long-term period into several time slots, and the spatial balance algorithm achieves energy optimization by balancing user requests in each time slot. Besides, a research case is simulated by Cloudsim platform. We conduct several comparative analysis in aspect of energy utilization, electricity cost and carbon emission. The experiment results show that the dispatcher can effectively reduce the electricity cost, and reasonably achieve the request distribution balance in geo-distributed hybrid powered data centers. [ABSTRACT FROM AUTHOR]

Emotion analysis, an integral aspect of human–machine interactions, has witnessed significant advancements in recent years. With the rise of multimodal data sources such as speech, text, and images, there is a profound need for a comprehensive review of pivotal elements within this domain. Our paper delves deep into the realm of emotion analysis, examining multimodal data sources encompassing speech, text, images, and physiological signals. We provide a curated overview of relevant literature, academic forums, and competitions. Emphasis is laid on dissecting unimodal processing methods, including preprocessing, feature extraction, and tools across speech, text, images, and physiological signals. We further discuss the nuances of multimodal data fusion techniques, spotlighting early, late, model, and hybrid fusion strategies. Key findings indicate the essentiality of analyzing emotions across multiple modalities. Detailed discussions on emotion elicitation, expression, and representation models are presented. Moreover, we uncover challenges such as dataset creation, modality synchronization, model efficiency, limited data scenarios, cross-domain applicability, and the handling of missing modalities. Practical solutions and suggestions are provided to address these challenges. The realm of multimodal emotion analysis is vast, with numerous applications ranging from driver sentiment detection to medical evaluations. Our comprehensive review serves as a valuable resource for both scholars and industry professionals. It not only sheds light on the current state of research but also highlights potential directions for future innovations. The insights garnered from this paper are expected to pave the way for subsequent advancements in deep multimodal emotion analysis tailored for real-world deployments. [ABSTRACT FROM AUTHOR]

Conclusion: This paper proposes a novel dual-pointer structure with a root node (called L2SM) and designs three optimization algorithms to improve the performance of L2SM. These improvements greatly enhance point and range query performance and reduce memory overhead without significant write performance loss. [ABSTRACT FROM AUTHOR]

Cloud-native database systems have started to gain broad support and popularity due to more and more applications and systems moving to the cloud. Various cloud-native databases have been emerging in recent years, but their developments are still in the primary stage. At this stage, database developers are generally confused about improving the performance of the database by applying AI technologies. The maturity model can help database developers formulate the measures and clarify the improvement path during development. However, the current maturity models are unsuitable for cloud-native databases since their architecture and resource management differ from traditional databases. Hence, we propose a maturity model for AI-empowered cloud-native databases from the perspective of resource management. We employ a systematic literature review and expert interviews to conduct the maturity model. Also, we develop an assessment tool based on the maturity model to help developers assess cloud-native databases. And we provide an assessment case to prove our maturity model. The assessment case results show that the database's development direction conforms to the maturity model. It proves the effectiveness of the maturity model. [ABSTRACT FROM AUTHOR]