Denoising diffusion probabilistic models that were initially proposed for realistic image generation have recently shown success in various perception tasks (e.g., object detection and image segmentation) and are incr...
Denoising diffusion probabilistic models that were initially proposed for realistic image generation have recently shown success in various perception tasks (e.g., object detection and image segmentation) and are increasingly gaining attention in computer vision. However, extending such models to multi-frame human pose estimation is non-trivial due to the presence of the additional temporal dimension in videos. More importantly, learning representations that focus on keypoint regions is crucial for accurate localization of human joints. Nevertheless, the adaptation of the diffusion-based methods remains unclear on how to achieve such objective. In this paper, we present DiffPose, a novel diffusion architecture that formulates video-based human pose estimation as a conditional heatmap generation problem. First, to better leverage temporal information, we propose SpatioTemporal Representation Learner which aggregates visual evidences across frames and uses the resulting features in each denoising step as a condition. In addition, we present a mechanism called Lookup-based Multi-Scale Feature Interaction that determines the correlations between local joints and global contexts across multiple scales. This mechanism generates delicate representations that focus on keypoint regions. Altogether, by extending diffusion models, we show two unique characteristics from DiffPose on pose estimation task: (i) the ability to combine multiple sets of pose estimates to improve prediction accuracy, particularly for challenging joints, and (ii) the ability to adjust the number of iterative steps for feature refinement without retraining the model. DiffPose sets new state-of-the-art results on three benchmarks: PoseTrack2017, PoseTrack2018, and PoseTrack21.
Objective:The annual influenza epidemic is a heavy burden on the health care system,and has increasingly become a major public health problem in some areas,such as Hong Kong(China).Therefore,based on a variety of mach...
详细信息
Objective:The annual influenza epidemic is a heavy burden on the health care system,and has increasingly become a major public health problem in some areas,such as Hong Kong(China).Therefore,based on a variety of machine learning methods,and considering the seasonal influenza in Hong Kong,the study aims to establish a Combinatorial Judgment Classifier(CJC)model to classify the epidemic trend and improve the accuracy of influenza epidemic early warning.
Dynamic graphs have emerged as a pivotal data structure underpinning real-world network applications. Against this backdrop, detecting anomalies in dynamic graphs has become particularly important, serving as a founda...
详细信息
Kidney tumor is a health concern that affects kidney cells and may leads to mortality depending on their type. Benign tumors can be unproblematic whereas malignant tumors pose the threat of kidney cancer. Early detect...
Kidney tumor is a health concern that affects kidney cells and may leads to mortality depending on their type. Benign tumors can be unproblematic whereas malignant tumors pose the threat of kidney cancer. Early detection and diagnosis are possible through kidney tumor recognition based on deep learning techniques. In this paper, a method based on transfer learning using deep convolutional neural network (DCNN) is proposed to recognize kidney tumor from computed tomography (CT) images. The proposed method was evaluated on 5284 images. The final accuracy, precision, recall, specificity and F1 score were 92.54%, 80.45%, 93.02%, 92.38% and 0.8628, respectively.
This paper focuses on the development of Complementary metal-oxide semiconductor(CMOS) image sensor and its applications in aerospace,medical and automotive ***,the representative events in history and the contributio...
This paper focuses on the development of Complementary metal-oxide semiconductor(CMOS) image sensor and its applications in aerospace,medical and automotive ***,the representative events in history and the contributions of some companies to CMOS image sensor are ***,some characteristics of CMOS image sensor are analyzed in the image field *** order to evaluate the performance of CMOS image sensor,single even effect and electronic endoscope structures are analyzed and active and passive range finder experiments are carried *** results show that the imaging based on CMOS sensor can fully meet the requirements of imaging applications in many fields.
This paper introduces an innovative data-driven approach for replicating behaviors in interconnected and heterogeneous dynamic systems. The core concept involves real-time control of dynamic systems to closely mimic r...
This paper introduces an innovative data-driven approach for replicating behaviors in interconnected and heterogeneous dynamic systems. The core concept involves real-time control of dynamic systems to closely mimic reference-model trajectories using model-free techniques. Within this coupled framework, one component possesses complete information about reference-trajectories, although not necessarily their dynamics. In contrast, follower systems, with limited connectivity to reference-model trajectories, exclusively replicate the behavior of the primary process, which retains insight into model-reference dynamics. The adopted strategies are causal, integrating higher-order error dynamics to ensure precise tracking of reference-trajectories. Furthermore, these strategies incorporate variations in reference-model dynamics via a pseudo partial derivative, akin to sensitivity derivatives in model-reference adaptive strategies. To optimize the dynamic behavior of the follower process, the solution employs a reinforcement learning mechanism through adaptive critics. This mechanism approximates the optimal strategy and the associated value function. The actor and critic weights of the adaptive critic structure are tuned using a projection technique to ensure convergence of the adapted strategy. The validation of this solution is demonstrated on a dynamic system with delays, simulating an underwater vehicle scenario. The developed methodology is rigorously compared with another high-order model-free adaptive control approach. The presented approach showcases its capability to effectively replicate behaviors, resulting in improved tracking accuracy.
Blockchain is a decentralized distributed ledger database. Consensus protocol is the core protocol of blockchain to solve Byzantine agreement problem. To let all blockchain nodes reach an agreement, the most commonly ...
详细信息
Bio-inspired soft robots present distinctive superiorities in safety issues working in a human-centered environment. Soft robotic hands are of prominent popularity for soft robots to be applied in real applications. W...
详细信息
After three years of multiple waves, COVID-19 has become epidemic, causing recurrent outbreaks. Many of COVID-19 cases have mild symptoms self-assessed at home, making it difficult to acquire formal laboratory data. T...
详细信息
Conformance checking compares a process model to its corresponding execution log, to detect inconsistencies and improve compliance with business processes. Nowadays, driven by trends such as big data and process autom...
详细信息
暂无评论