12 hasil untuk "QING Lin-Bo"
Menampilkan 1–12 · 0,45 detik · RSS
-
Artikel Jurnal
Emotion recognition of the natural scenes based on attention mechanism and multi-scale network
Emotion recognition as a fundamental topic in computer vision has made tremendous progress, yet emotion recognition in unconstrained environments is still challenging. Existing methods mainly use face, posture, and scene information to recognize emotions, but these methods ignore the uncertainty of…
-
Artikel Jurnal
An Improved Moving Target Tracking Method Suitable for Edge Computing Framework
With the rapid development of 5G communications technology, IoT and big data technology, traditional cloud computing models have become increasingly unable to keep up with the growth rate of data, as a new computing model, edge computing has demonstrated a strong ability to handle big data and…
-
Artikel Jurnal
Characterization of the Pore Morphology of 3D Rock Models Using Spherical Harmonic Functions
Aiming at the statistical characteristic function and conventional spherical harmonic functions could not effectively describe 3D rock models, rock spherical harmonic functions were proposed. This algorithm was an improvement on the conventional spherical harmonic functions used in 3D model…
-
Artikel Jurnal
Facial pain assessment based on contrastive learning
Automatic pain assessment has great clinical application prospects and has attracted great attention, because it is feasible to recognize facial pain through facial visual information.Existing methods mostly focus on extracting features of different modalities and different scales, without…
-
Artikel Jurnal
Emotion recognition based on fusion of expression and posture
Research shows that the role of body posture in emotion recognition is always underestimated. Aiming at the problems of low face resolution and low expression recognition accuracy in public space, an emotion recognition method based on facial expression and body posture is proposed. Firstly, the…
-
Artikel Jurnal
Remote sensing image scene classification based on multi-channel self-attention network
High resolution remote sensing image (HRRSI) scene classification is widely used in many fields such as land monitoring, environment protection, urban planning and so on. The existing scene classification methods cannot fuse the local-texture and globalsemantic information well, and the…
-
Artikel Jurnal
A novel error resilience scheme for key frames in Wyner-Ziv video coding
In Wyner-Ziv video coding system, correct transimission and accurate decoding of key frames would affect the whole system performance directly. However, in wireless communication system, data loss of key frames often happened due to network failure, network congestion or other reasons. Data lossy…
-
Artikel Jurnal
Multi person behavior recognition based on scene and interactive features
Human behavior is complex and diverse, and the information such as scene, appearance and location are closely related to human behavior. Aiming at the problem of how to make efficient comprehensive use of these information, a multi-person behavior recognition method integrating scene and…
-
Artikel Jurnal
Analysis of the impact of environmental elements on inclusiveness and resident attribute preferences based on street view images
Inclusiveness refers to the openness and accessibility of urban space to different attribute groups, while preferences refer to the preferences and selections of different resident attribute groups for environmental elements, which together reflect the adaptability and satisfaction of cities to…
-
Artikel Jurnal
Fast Stitching Algorithm Based on Perceptual HASH and Scale Invariant Feature Transformation
The k-d tree registration algorithm has the disadvantages of too much computation for high-dimensional features, less use of SIFT feature information and a lot of errors in the registration results, due to the complexity of scale invariant feature transformation (SIFT). Therefore, a fast stitching…
-
Artikel Jurnal Teks lengkap
Text-to-Image Generation Technology Based on Transformer Cross Attention
In recent years,the research on the methods of text to image based on generative adversarial network (GAN) continues to grow in popularity and have made some progress.The key of text-to-image generation technology is to build a bridge between the text information and the visual information,and…
-
Artikel Jurnal
Image emotion distribution learning based on multiscale feature fusion
Visual emotion analysis aims to analyze the emotional response of human beings to visual stimuli, which has attracted multimedia visual data related fields such as sharing platforms and social networking in recent years. Traditional image emotion analysis focuses on the classification of single…