Facial expression recognition plays a key role in promoting the development of comprehensive intelligence and building friendly human-computer interaction. Due to the interference of feature noise in expression data, ...
详细信息
ISBN:
(数字)9798331521950
ISBN:
(纸本)9798331521967
Facial expression recognition plays a key role in promoting the development of comprehensive intelligence and building friendly human-computer interaction. Due to the interference of feature noise in expression data, the lightweight facial expression recognition model with fewer parameters is difficult to learn more expression features through simple training, which limits the improvement of its recognition performance. An efficient facial expression recognition network based on Spot-adaptive Knowledge Distillation is proposed in this paper. Inspired by VoVNetV2, the network designed in this paper is lightweightly improved using Depthwise Separable Convolution and the parameter-free SimAM attention mechanism, reducing the number of parameters to 0.21 M. To further improve the recognition accuracy of the model, Spot-adaptive Knowledge Distillation is employed to improve the characterization ability of the model. The recognition accuracies of the student network designed in this paper on the KDEF and RAF-DB datasets are 93.05% and 81.17% respectively after spot-adaptive distillation.
作者:
Nanxin HuangChi XuSchool of Automation
China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China
Driven by advancements in industrial production and artificial intelligence, the need for pose estimation of new ob-jects in areas like robotic manipulation and virtual reality is increasing. We introduce a zero-shot ...
详细信息
ISBN:
(数字)9798331521950
ISBN:
(纸本)9798331521967
Driven by advancements in industrial production and artificial intelligence, the need for pose estimation of new ob-jects in areas like robotic manipulation and virtual reality is increasing. We introduce a zero-shot object pose estimation approach that identifies the poses of objects excluded from the training dataset, removing the requirement for re-modeling. The method is built around a multi-level features fusion framework de-signed to enhance generalization. First, a trainable feature extraction module filters and selects multi-level features extracted by the backbone network. Unlike traditional convolutional ker-nels, we incorporate a dynamic convolution kernel to enhance the feature extraction capability. Second, in the feature fusion module, we adopt a dynamic weight generation strategy to perform weighted fusion of multi-level features. This method enhances template matching by effectively describing similarities between unseen objects (those absent from the training set) and templates, leveraging robust and adaptive feature representations to narrow the gap with seen objects. Experimental results demonstrate that our approach achieves state-of-the-art performance on two popu-lar benchmark datasets, LineMod and LineMod-Occlusion, proves that our method has better generalization than previous models.
The fluxgate sensor is the most widely used sensor in vector magnetic measurement. However, during long-term continuous observation, the fluxgate sensor will produce large measurement errors due to changes in ambient ...
详细信息
A single-modal infrared or visible image offers limited representation in scenes with lighting degradation or extreme weather. We propose a multi-modal fusion framework, named SDSFusion, for all-day and all-weather in...
详细信息
Due to the rapid growth of online education worldwide, assessing the learning effectiveness of students during online classes has become increasingly challenging for teachers. In this paper, a method of assessing onli...
详细信息
In landslide susceptibility prediction using machine learning (ML) models, the selection of appropriate negative samples significantly impacts the accuracy of the predictions. Many existing studies tend to randomly ch...
详细信息
Establishing the dynamics model of the offshore drilling experimental system can better complete the offshore drilling test in the laboratory environment and reduce the cost of testing. A dynamical modeling method for...
详细信息
In this paper, the preassigned-time synchronization (PTS) problem for a fifth-order memristive chaotic circuit (MCC) is investigated by designing a time-dependent intermittent controller. First, the dynamic characteri...
详细信息
This paper proposes an inverse compensation feed-forward control strategy for a vertical pneumatic artificial muscle (PAM) system. Firstly, we conduct a preliminary experiment on the vertical PAM system, and on this b...
详细信息
This paper is concerned with group consensus of multi-agent systems(MASs) that consist of two groups in additive noise environments. First, a control protocol is proposed based on the state information of each agent&#...
详细信息
This paper is concerned with group consensus of multi-agent systems(MASs) that consist of two groups in additive noise environments. First, a control protocol is proposed based on the state information of each agent's neighbors corrupted by additive noises. Second, some sufficient conditions and necessary conditions are obtained for the following two types of group consensus behaviors.(1) Pure group consensus:agents in both groups have the same behavior(weak consensus or strong consensus);(2) hybrid group consensus: agents in different groups achieve different consensus behaviors. It is revealed that the influence between the two groups should be attenuated such that the MASs can achieve group consensus in additive noise environments, i.e., the affected group must fight against the influence that comes from another ***, some simulation examples are given to illustrate the feasibility of the theoretical results.
暂无评论