A Few-Shot Steel Surface Defect Generation Method Based on Diffusion Models by Hongjie Li, Yang Liu, Chuni Liu, Hongxuan Pang, Ke Xu

“…Furthermore, when the generated samples were used for data augmentation, the detection network performance was effectively enhanced. Detection recall significantly improved for both defect classes, increasing from 0.656 to 0.908 for Inclusion defects and from 0.86 to 0.986 for Foreign Object Embedding defects. …”
Get full text

Screen shooting resistant watermarking based on cross attention by Lianshan Liu, Peng Xu, Qianwen Xue

“…Most existing solutions are based on Convolutional Neural Networks (CNNs) for the embedding of watermarks. However, due to the limited reception field of CNNs, they are proficient in extracting local features but cannot understand the entire image. …”
Get full text

Research on Early Fault Identification of Cables Based on the Fusion of MTF-GAF and Multi-Head Attention Mechanism Features by Hao Wu, Dan Tang, Yuan Cai, Chaowen Zheng

“…At the same time, the classification results of cable early faults are visualized using the t-distributed Stochastic Neighbor Embedding (t-SNE) method to visually observe the classification effect of the hybrid neural network. …”
Get full text

Multimodal emotion recognition method in complex dynamic scenes by Long Liu, Qingquan Luo, Wenbo Zhang, Mengxuan Zhang, Bowen Zhai

“…Second, we introduce a dynamic-static cross fusion network (D-SCFN) to enhance the integration and extraction of dynamic and static information, embedding it seamlessly within the T5 framework. …”
Get full text

Automatic Mushroom Species Classification Model for Foodborne Disease Prevention Based on Vision Transformer by Boyuan Wang

“…We visualized the high-dimensional outputs of the ViT-L/32 model to achieve the interpretability of ViT-L/32 using the t-distributed stochastic neighbor embedding (t-SNE) method. …”
Get full text

Multi-Scale Contrastive Learning with Hierarchical Knowledge Synergy for Visible-Infrared Person Re-Identification by Yongheng Qian, Su-Kit Tang

“…Most existing works focus on learning shared feature representations from the final embedding space of advanced networks to alleviate modality differences between visible and infrared images. …”
Get full text

Huffman coding-based data reduction and quadristego logic for secure image steganography by Irsyad Fikriansyah Ramadhan, Ntivuguruzwa Jean De La Croix, Tohari Ahmad, Andre Uzamurengera

“…Furthermore, current methods lack adaptability to diverse cover media and struggle to maintain reversibility and high visual quality under increased embedding capacities. …”
Get full text

Art and Language After AI by Anil Bawa-Cavia

Get full text

An automated construction method of 3D knowledge graph based on multi-agent systems in virtual geographic scene by Yukun Guo, Jun Zhu, Jianlin Wu, Jinbin Zhang, Zhihao Guo, Weilian Li, Heng Zhang

“…Virtual geographic scenes can significantly facilitate the visualization and comprehension of the real world. However, the multitude of object types and complex relationships they contain make it challenging to display embedded geographic knowledge. …”
Get full text

An Improved Fault Diagnosis Method and Its Application in Compound Fault Diagnosis for Paper Delivery Structure Coupling by Fu Liu, Haopeng Chen, Yan Wang

“…Furthermore, the dimensionality reduction results of each network layer are visualized using the T-stochastic neighbor embedding (T-SNE) method, which reveals clear feature patterns and confirms the model’s reliability and effectiveness.…”
Get full text

An Improved DeepSORT-Based Model for Multi-Target Tracking of Underwater Fish by Shengnan Liu, Jiapeng Zhang, Haojun Zheng, Cheng Qian, Shijing Liu

“…This study proposes an underwater fish object tracking method based on the improved DeepSORT algorithm, utilizing ResNet as the backbone network, embedding Deformable Convolutional Networks v2 to enhance adaptive receptive field capabilities, introducing Triplet Loss function to improve discrimination ability among similar fish, and integrating Convolutional Block Attention Module to enhance key feature learning. …”
Get full text

Robust circulating microRNA signature for the diagnosis and early detection of pancreatobiliary cancer by Shuichi Mitsunaga, Masafumi Ikeda, Makoto Ueno, Satoshi Kobayashi, Masahiro Tsuda, Ikuya Miki, Takamichi Kuwahara, Kazuo Hara, Yukiko Takayama, Yutaro Matsunaga, Keiji Hanada, Akinori Shimizu, Hitoshi Yoshida, Tomohiro Nomoto, Kenji Takahashi, Hidetaka Iwamoto, Hideaki Iwama, Etsuro Hatano, Kohei Nakata, Masafumi Nakamura, Hiroko Sudo, Satoko Takizawa, Atsushi Ochiai

“…The optimized serum processing conditions were evaluated using t-distributed stochastic neighbor embedding (t-SNE) visualization. Serum miRNA candidates for disease association were selected using weighted gene coexpression network analysis (WGCNA). …”
Get full text

Reversible data hiding scheme based on enhanced image smoothness by Jinghan WANG, Hui ZHU, Helin LI, Hui LI, Xiaopeng YANG

“…With the prosperity of Internet technology and the popularity of social networks, reversible data hiding technology has been widely adopted in concealed information transmission of medical and military fields with its advantages on secret information recovery.Traditional reversible data hiding schemes mainly focus on the enhancement of embedding capacity and the reduction of the distortion rate of stego image, but pay less attention to the understanding of image details with the human eyes.Thus, it is difficult to resist hidden information detection methods.To solve the above challenge, a reversible data hiding algorithm was proposed, which ensured the visual quality of the stego image in the process of data hiding through the image visual smoothness enhancement.Specifically, the original image was divided into reference area and non-reference area.The secret data was embedded through the translation of the difference, which was calculated according to the predicted pixel value and the original pixel value of the non-reference area.To guarantee the visual quality of the image, smoothing mechanism was constructed, in which a Gaussian filter was utilized as a template to filter the predicted value and to add the filter difference into the cover image without loss.The pixel value of the reference region was used as edge information for lossless restoration of the original image.The filtering coefficient in Gaussian function was exploited as the embedded key to ensure the security of secret information.Simulation results regarding a large number of classical image data sets illustrated that the visual smoothness of stego image processed by this scheme was effectively enhanced with lower distortion rate, higher embedding rate, and higher embedding and extraction efficiency.In a typical circumstance, the similarity between the generated stego image and the Gaussian filter image can reach 0.9963.The PSNR and the embedded capacity can be up to 37.346 and 0.3289 BPP, respectively.…”
Get full text

A computational lens into how music characterizes genre in film. by Benjamin Ma, Timothy Greer, Dillon Knox, Shrikanth Narayanan

“…We construct supervised neural network models with various pooling mechanisms to predict a film's genre from its soundtrack. …”
Get full text

4KSecure: A Universal Method for Active Manipulation Detection in Images of Any Resolution by Paweł Duszejko, Zbigniew Piotrowski

“…This paper presents a method for actively safeguarding image integrity based on embedding a hidden signature generated by a neural network. …”
Get full text

Identification of diabetic retinopathy lesions in fundus images by integrating CNN and vision mamba models. by Zenglei Liu, Ailian Gao, Hui Sheng, Xueling Wang

“…Furthermore, the vision mamba component incorporates the bidirectional state space method and positional embedding to enable the location sensitivity of visual data samples and meet the conditions for global relationship context. …”
Get full text

The CLIP - GPT Image Captioning Model Integrated with Global Semantics by TAO Rui, REN Honge, CAO Haiyan

“…In this paper, we introduce a new method that breaks through the limitation of visual feature classification by dividing images into patches as visual semantic units for open-vocabulary cross-modal association with language features. …”
Get full text

Adaptive Coati Optimization Enabled Deep CNN-based Image Captioning by Balasubramaniam S, Seifedine Kadry, Rajesh Kumar Dhanaraj, Satheesh Kumar K

“…The extracted features and detected objects are given to image captioning which is exploited by Deep Convolutional Neural Network (Deep CNN). The Deep CNN is trained by using the proposed Adaptive Coati Optimization Algorithm (ACOA). …”
Get full text

Providing context: Extracting non-linear and dynamic temporal motifs from brain activity. by Eloy Geenjaar, Donghyun Kim, Vince Calhoun

“…Interestingly, patients appear to spend more time in three clusters, one closer to controls which shows increased visual-sensorimotor, cerebellar-subcortical, and reduced cerebellar-visual functional network connectivity (FNC), an intermediate station showing increased subcortical-sensorimotor FNC, and one that shows decreased visual-sensorimotor, decreased subcortical-sensorimotor, and increased visual-subcortical domains. …”
Get full text

Replicating associative learning of rodents with a neuromorphic robot in an open-field arena by Tianze Liu, Kang Jun Bai, Hongyu An

“…Different coding schemes—rate coding for vibration signals and population coding for visual signals—were implemented. The associative learning model employs 19 spiking neurons and follows Hebbian plasticity principles to associate visual cues with favorable or unfavorable locations. …”
Get full text