International Journal of Pattern Recognition and Artificial Intelligence

Exploring proper way to conduct multi-speech feature fusion for cross-corpus speech emotion recognition is crucial as different audio features could provide complementary cues reflecting human emotion status. Speech emotion recognition allows computers to analyze the specific emotional condition of the speaker through speech, which is of great significance to the development of human–computer interaction technology. While most previous approaches only extract a single speech feature for emotion recognition, existing fusion methods such as concatenation, parallel connection, and splicing ignore heterogeneous patterns in the interaction between features and features, resulting in performance of existing systems. In this paper, we propose a novel graph-based fusion method to explicitly model the relationships between every pair of audio features, which provides a new research idea for speech feature fusion. Specifically, we propose a multi-dimensional edge features learning strategy called graph-based multi-feature fusion method for speech emotion recognition. It represents each speech feature as a node and learns multi-dimensional edge features to explicitly describe the relationship between each feature-feature pair in the context of emotion recognition. This way, the learned multi-dimensional edge features encode speech feature-level information from both the vertex and edge dimensions. Our approach consists of three modules: an Audio Feature Generation (AFG) module, an Audio-Feature Multi-dimensional Edge Feature (AMEF) module and a Speech Emotion Recognition (SER) module. The proposed methodology yielded satisfactory outcomes on the SEWA dataset. Furthermore, the method demonstrated enhanced performance compared to the baseline in the AVEC 2019 Workshop and Challenge. We used data from two cultures as our training and validation sets: two cultures containing German and Hungarian on the SEWA dataset, the CCC scores for German are improved by 17.28% for arousal and 7.93% for liking, and for Hungarian, the CCC scores are improved by 11.15% for arousal and 131.11% for valence. The outcomes of our methodology demonstrate a 13% improvement over alternative fusion techniques, including those employing one-dimensional edge-based feature fusion approach. The experiments on some parts of the Aff-Wild 2 dataset demonstrate that our approach exhibits a certain degree of generalizability and robustness. Code is available at https://github.com/ChaosWang666/Graph-based-multi-Feature-fusion-method.

Floorplanning is a pivotal step in the early stages of integrated circuit physical design. In modern design, the outline of the chip is typically determined before floorplanning and each module must be placed within the specified fixed outline. However, when dealing with soft modules, the task becomes even more complex and challenging as it requires simultaneous determination of both position and shape. This paper proposes a genetic algorithm based on random subtree crossover to address the floorplanning problem of soft modules under fixed-outline constraints. First, a slicing tree is employed to represent the floorplan solution due to its advantage in handling the flexibility of soft modules. A three-step random subtree crossover (RSC) is proposed to generate feasible descendants specifically for slicing tree representation. Second, we adopt three types of perturbations as mutation operators and determine the probability of selecting a mutation operator on account of its empirical performance to improve the effectiveness of mutation. Then, to enhance the genetic algorithm’s rate of convergence, a step named insert after deletion (IAD) is incorporated into the evolutionary process to eliminate operator nodes with large dead spaces. Finally, the algorithm is verified by employing two standard circuits, MCNC and GSRC. The test consequences show that the proposed algorithm realized a 100% success rate across various aspect ratios while surpassing previous works in terms of wire length performance.

This paper presents AMFANet, an advanced deep learning model engineered for high-quality image style transfer. AMFANet integrates cutting-edge techniques such as the Adaptive Multi-Scale Feature Fusion (AMSF) module and Hybrid Attention Mechanism (HAM) to significantly improve style consistency, content fidelity, and texture preservation. The model also utilizes Segmented Atrous Spatial Pyramid Pooling (SASPP) for effective multi-scale feature extraction. Comprehensive experimental evaluations demonstrate that AMFANet surpasses current state-of-the-art models like StyleGAN3, ChipGAN, ACL-GAN, and CycleGAN in generating high-fidelity stylized images while preserving intricate details and artistic essence. Future research will focus on optimizing computational efficiency, enabling multi-style transfer, enhancing user interaction, and exploring cross-domain applications. These findings highlight AMFANet’s potential as a robust solution for advanced image style transfer in both artistic and practical domains.

Sentiment analysis using scene text images is complex and challenging because it has an arbitrary background, and the method should rely on only visual features. Unlike most existing methods that use either text or images or both, this study uses only scene text images for sentiment analysis. The intuition to use only scene text images is that sometimes users express their feelings and emotions or convey their messages by writing text in different shapes with diverse background designs. It is noted that the existing methods ignore such vital cues for sentiment analysis. This work explores a vision transformer to extract visual features that represent contextual information about the appearance of the text image. Further, to strengthen the visual features, the proposed work introduces contrastive learning which maximizes the gap between inter-classes and minimizes the gap between intra-classes of positive, negative, and neutral. To demonstrate the effectiveness of the proposed method, it is tested on our own constructed dataset and benchmark dataset. A comparative study of our method with the existing method shows the proposed method is superior in the classification of positive, negative, and neutral scene text images.

The rapid development of big data and artificial intelligence technology has given new vitality and value to smart cities, which provide rich algorithm models and knowledge computing capabilities. Since the recognition results of the traditional Convolutional Neural Network (CNN) model in the face database are prone to over-fitting, this paper proposes a face recognition algorithm based on an improved CNN model and deep learning. The improved CNN model has a good network image recognition performance, improves the data training speed, and optimizes the network structure parameters. Based on an improved CNN model, using the facial expression recognition (FER2013) data to test the model performance, to achieve more accurate face recognition. Experimental results show that the recognition rate of the improved model and deep learning algorithm on the FER2013 dataset reaches 98.36%, which is 4.63% higher than before the improvement.

Segment Anything Model (SAM), an advanced universal image segmentation model trained on an expansive visual dataset, has set a new benchmark in image segmentation and computer vision. However, it faced challenges when it came to distinguishing between shadows and their backgrounds. To address this, we proposed ShadClips, which consists of SAM-optimizer and SONet. It has dramatically enhanced SAM’s ability to segment shadow images, differentiating between the background and both soft and hard shadows adeptly. Due to its dependence on pixel point inputs, the SAM-Optimizer interface could do better. This method presents challenges, especially when dealing with long, extended shadows. To make the user experience more intuitive and effective, we incorporated the capabilities of CLIPs. Therefore, simple text descriptions like “A photo of a shadow” can be used to guide the SAM-Optimizer, allowing it to select the most relevant shadow mask from SAM’s comprehensive category list. Meanwhile, we introduce SONet to shadow removal. A large number of experiments on ISTD/SRD prove that the proposed method is effective and satisfactory. The source code of the ShadClips can be accessed from https://github.com/zhangbaijin/SAM-helps-Shadow.

Regarding global path planning for search and rescue robots, the traditional method of A* algorithm is slow and has many turning points along the intended route, which hinders its searching speed. A new and enhanced A* algorithm is presented in this paper, incorporating the Floyd-trajectory optimization algorithm. The search directions of the traditional A* algorithm were refined from eight to five. Next, we improved the cost estimation function by introducing a weight value to balance the estimated heuristic function value with the actual cost value, thereby enhancing the algorithm’s search efficiency. Finally, we applied the Floyd path optimization algorithm to eliminate redundant nodes from the route. The use of both the Floyd and A* algorithms in combination has been proven through simulations and experiments to improve search results by 40% compared to solely using the A* algorithm. This integration effectively reduces the search scope, diminishes the number of turning points by 63.8%, improves path smoothness, and effectively shortens the planned path length.

To address the limitations of visible light cameras that cannot function effectively at night, infrared cameras have become the optimal supplement. However, current methods for visible–infrared cross-modality person re-identification focus solely on feature combination and fusion, neglecting the importance of feature alignment. To address this issue, we introduce a novel Hierarchical Feature Fusion (HFF) network, which comprehensively integrates features across various levels through sequential feature extraction. Specifically, we design a pixel-level contrastive loss function that makes pixels in the same region of cross-modality images more similar and distinguishes pixel features at different locations, thereby extracting similar low-frequency information in the shallow network. Furthermore, in the deep network, we extract high-frequency information of different modalities through the Bi-Transformer Layer and propose Node-level Coupling Attention and Modality-level Decoupling Attention. Coupling attention is used for high-frequency information coupling within the same modality while decoupling attention is used for high-frequency information decoupling between different modalities to obtain more texture and detail information. Through a series of experimental results, we validate the superiority of the proposed HFF network in cross-modality person re-identification. Our proposed method achieved 87.16% and 95.23% Rank-1 on the SYSU-MM01 and RegDB datasets, respectively, and extensive experiments have validated its effectiveness in feature alignment.

To improve the real-time performance and the target adaptability of penetration fuze detonation control systems, and to enhance the system fusion processing capability for multi-sensor information, this paper uses a modular design concept to construct a miniaturized (ø38mm×4mm) fuze detonation control system that is capable of real-time processing of data from multiple information sources. The core component of this system is the GD32E230 microcontroller, which features a high dominant frequency and low power consumption. This device is integrated with a ferroelectric memory and signal processing circuits that match the sensors. To address the issue of unclear traditional acceleration signal penetration and the difficulties associated with the identification of these signals, the approach in this paper improves feature recognition accuracy through rapid acquisition and fusion of multiple types of sensor output signal, and self-adaptive identification of multilayered targets and single-layer thick targets is achieved. During the programming of the embedded system, the hardware register is operated directly, the instruction execution sequence is optimized, and the program execution efficiency is improved by using the function characteristic that some microcontroller unit peripherals do not occupy the central processing unit when working, thus allowing the intended purpose of improving the system’s real-time performance to be achieved. A semi-physical simulation method is then used to verify the performance of the penetration fuze detonation control system. The results obtained show that the system has 100%-layer counting accuracy for multilayered targets and a relative error of less than 1% for the calculated residual velocities of single-layer thick targets, thus validating the effectiveness of the system.

A novel parameter optimization method, named the geyser algorithm based on particle swarm optimization (GEA–PSO) method, was proposed for optimizing the key parameters of a bird-like flapping wing robot. The kinematic model, the multiple objective functions, and some constraint conditions were first established. A novel GEA–PSO algorithm was designed. Some test functions verified the good superiority of the GEA–PSO method with a fast velocity and high accuracy. The key parameters of the robot were successfully optimized using the GEA–PSO method. Kinematic simulations were carried out to obtain some key quantities, such as the end trajectory, the flapping angle, and the folding angle. Robot experiments were conducted based on the robot prototype, which verified the effectiveness of the optimization method. The GEA–PSO method plays a crucial role in the parameter optimizations and the stable flights of the robot.