Search

Copyright of Journal of Beijing University of Technology is the property of Journal of Beijing University of Technology, Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

As the foundation for implementing closed-loop detection within the realm of visual SLAM (simultaneous localization and mapping), visual place recognition (VPR) has great potential in various applications of greenhouse robot navigation and other fields. However, the existing VPR cannot fully meet the actual requirements of greenhouse scenes due to the complexity and constant variations in the greenhouse environment. In particular, the local feature aggregation paradigm strongly depends on the induction bias of training samples in VPR models, which leads to the issue of information redundancy during feature aggregation. In this study, a greenhouse VPR was presented, according to the optimal transport of local feature aggregation. The process of aggregating local features into a global descriptor was framed as an optimal transport problem, where the cost matrix was predicted through an MLP (multi-layer perceptron). Thus, a cost matrix was dynamically generated using the local features that was extracted from the greenhouse scene images. Additionally, a 'dustbin' cluster was introduced into the cost matrix to allocate the redundant features. Taking the cost matrix as the input, the Sinkhorn algorithm was employed to determine an optimal solution to the assignment matrix. Furthermore, the soft assignment of local features to various clusters was achieved through the assignment matrix. Ultimately, the assignment was concatenated to form a global descriptor for the scene image, which was used for place recognition. A deep neural network (DNN) was optimized and designed to serve as the backbone for local feature extraction of greenhouse scene images, by combining the advantages of CNN (convolutional neural network) and Transformer. Furthermore, cosine similarity was used as the metric function to calculate the similarity measure between scene image global descriptors, so as to perform descriptor matching. A series of experiments were conducted in a tomato greenhouse. The experimental results showed that the improved model achieved better performance. The top-1 recall rate (R@1) for place recognition was achieved at 88.96%, which was 29.67, 2.97, and 2.89 percentage points higher than the those of NetVLAD, MixVPR, and EigenPlaces models, respectively. When compared to the aggregators employed in MixVPR and NetVLAD, our aggregator achieved improvements in R@1 by 1.09 and 21.65 percentage points, respectively, showcasing its effectiveness. Compared with the CNN, the improved network achieved an increase of 5.45 percentage points in R@1. There was even more pronounced R@1 improvement (reaching 10.48 percentage points), compared with a Transformer network. Simultaneously, our network resulted in a 1.6-fold increase in computation speed compared to the previous Transformer. In addition, the experiments further demonstrated that the improved model exhibited excellent performance of place recognition and strong robustness when dealing with factors, such as small sampling distance shifts, small viewpoint shifts, and different sunlight intensities. The greenhouse VPR achieved a place recognition rate of no less than 81.94% in actual greenhouses, indicating its practical application potential. The method based on optimal transport of local feature aggregation and global descriptor generation was effective for place recognition, and the image local feature extraction network can boost the performance of place recognition. These findings can provide technical support to the visual systems of intelligent agricultural machinery in the greenhouse. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Chongqing University of Posts & Telecommunications (Natural Science Edition) is the property of Chongqing University of Posts & Telecommunications and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Human gaze behavior, as a non-verbal cue, plays a crucial role in revealing human intentions. Gaze target detection has attracted extensive attention from the machine vision community. However, existing gaze target detection methods usually focus on the texture information extraction of images, ignoring the importance of stereo depth information for gaze target detection, which makes it difficult to deal with scenes with complex texture. In this work, a novel gaze target detection network based on attention mechanism and depth prior is proposed, which adopts two-stage architecture (i.e., a gaze direction prediction stage and a saliency detection stage). In the gaze direction predication stage, a channel-spatial attention mechanism module is established to recalibrate texture features, and a head position encoding branch is designed to achieve texture and head position-aware enhanced high-representation features to accurately predict gaze. Furthermore, a strategy is proposed to introduce the depth representing the stereoscopic or distance information in the 3D scene as a prior into the saliency detection stage. At the same time, the channel-spatial attention mechanism is used to enhance the multi- scale texture features, and the advantages of depth geometric information and image texture information are fully utilized to improve the accuracy of gaze target detection. Experimental results show that the proposed model performs favorably against the state-of-the-art methods on GazeFollow and DLGaze datasets. [ABSTRACT FROM AUTHOR]

Aiming at the problems of local structure disconnection and inaccurate detail restoration in existing image restoration algorithms, this paper proposed an inpainting algorithm based on semantic prior and dual channel feature extraction（SPDCE）. The algorithm used semantic prior network to learn the semantic information and context knowledge of the missing area, predicted the missing area, and enhanced the local consistency of the generated image;then, the dual channel feature extraction network fully mined image information, enhanced the perception and utilization of texture details; next, the context feature adjustment module could capture and encode rich semantic features at multiple scales, thereby generating more realistic image views and finer texture details. After conducting experimental verification on the datasets CelebA-HQ and Places2, the results showed that compared with commonly used algorithms, the SPDCE algorithm improved the peak signal ratio（PSNR） and structural similarity（SSIM） by 1.6dB to 1.73dB and 3.1% to 9.9%, while L1 loss decreased by 15.2% to 27.8%. Experimental results show that the repaired image of the proposed algorithm has a more reasonable structure and richer details, and the inpainting effect is better. [ABSTRACT FROM AUTHOR]

For the problem of ATC speaker recognition, a method based on voiceprint feature aggregation is proposed, which could distinguish different speakers from an audio stream. First, we develop the ResNet spectrogram feature extractor and the NetVLAD feature fusion module, both of which seldom used in speaker recognition. Second, we insert two modules above and develop a novel end-to-end speaker recognition framework deriving from classic X-VECTORS method. Finally, the accuracy of the proposed method and the baseline method is compared and analyzed under a real ATC voice dataset. The results show that, compared with X-VECTORS network, the voiceprint aggregation method has superior recognition accuracy.

Deep learning-based image compressive sensing has received extensive attention due to its powerful learning ability and fast processing speed.With the increase in the depth of convolutional neural networks,the existing image reconstruction methods using neural networks do not fully utilize the residual features in the network.In order to solve this problem,this paper proposes a compressed sensing attention neural network framework based on residual feature aggregation(RFA2CSNet)by jointly optimizing the sampling and inverse reconstruction processes.First,the block compressed sensing sampling sub-network and the initial reconstruction sub-network are constructed to adaptively learn the measurement matrix and generate the initial reconstruction image.Then the residual learning and spatial attention mechanisms are introduced to construct the residual feature aggregation attention reconstruction sub-network to make the residual feature more focused on the key spatial content,so as to further improve the reconstructed image quality.Experimental results show that the proposed network is superior to the existing image compressed sensing reconstruction algorithm in the case of comparable reconstruction time,and obtains better image compressive sensing reconstruction quality.Specifically,using 11 images for testing with a sampling rate of 0.10,the average peak signal-to-noise ratio increases by 0.34~6.18 dB compared with other deep learning-based methods.

With the growth of Internet multimedia data,text image retrieval has become a research hotspot.In image and text retrieval,the mutual attention mechanism is used to achieve better image-text matching results by interacting image and text features.However,this method cannot obtain image features and text features separately,and requires interaction of image and text features in the later stage of large-scale retrieval,which consumes a lot of time and is not able to achieve fast retrieval and ma-tching.However,the cross-modal image text feature learning based on Transformer has achieved good results and has received more and more attention from researchers.This paper designs a novel Transformer-based text image retrieval network structure(HAS-Net),which mainly has the following improvements:a hierarchical Transformer coding structure is designed to better utilize the underlying grammatical information and high-level semantic information;the traditional global feature aggregation method is improved,and the self-attention mechanism is used to design a new feature aggregation method;by sharing the Transformer coding layer,image features and text features are mapped to a common feature coding space.Finally,experiments are conducted on the MS-COCO and Flickr30k datasets,the cross-modal retrieval performance has been improved,and it is in a leading position among similar algorithms.It is proved that the designed network structure is effective.

In order to solve the problem of reducing the accuracy of part segmentation due to the loss of secondary features by PointNet max pooling, this paper proposed attention weighted feature aggregation Point Net network for part segmentation, and it could make full use of different features of point cloud for part segmentation. Firstly, it used the multiple layer perceptions to extract the spatial geometric features of the point cloud, and then multiplied the features and the shared weight matrix to obtain the attention scores of each feature of each point. Then it used the normalized scores as the weight to multiply the corresponding global feature and summed to obtain the aggregated global feature. Finally, it used the multiple layer perceptions to map the aggregated feature to the part segmentation results. Experimental results show that the proposed method improves the overall average mean intersection over union of part segmentation than traditional PointNet, has significant advantages in network robustness and computational complexity, and effectively optimizes PointNet. [ABSTRACT FROM AUTHOR]

For the problem of ATC speaker recognition, a method based on voiceprint feature aggregation is proposed, which could distinguish different speakers from an audio stream. First, we develop the ResNet spectrogram feature extractor and the NetVLAD feature fusion module, both of which seldom used in speaker recognition. Second, we insert two modules above and develop a novel end-to-end speaker recognition framework deriving from classic X-VECTORS method. Finally, the accuracy of the proposed method and the baseline method is compared and analyzed under a real ATC voice dataset. The results show that, compared with X-VECTORS network, the voiceprint aggregation method has superior recognition accuracy. [ABSTRACT FROM AUTHOR]

Copyright of Chinese Journal of Liquid Crystal & Displays is the property of Chinese Journal of Liquid Crystal & Displays and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Guangxi Normal University - Natural Science Edition is the property of Gai Kan Bian Wei Hui and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of South China University of Technology (Natural Science Edition) is the property of South China University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer-Aided Design & Computer Graphics / Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao is the property of Gai Kan Bian Wei Hui and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In large ⁃scale road environment, point ⁃based methods require dynamic calculations, and voxel⁃ based methods often lose a lot of information when balancing resolution and performance. To overcome the drawbacks of the above two classical methods, this paper proposes a general network architecture that com⁃ bines bi ⁃level convolution and dynamic graph edge convolution optimization for multi ⁃object recognition of large ⁃ scale road scenes. The framework integrates the convolution operations of two different domains of points and supervoxels to avoid redundant calculations and storage of spatial information in the network. Coupled with the dynamic graph edge convolution optimization, our model enables it to process large⁃scale point clouds end⁃to⁃end at once. Our method was tested and evaluated on different datasets. The experimen⁃ tal results show that our method can achieve higher accuracy in complex road scenes, which is superior to the existing advanced methods. [ABSTRACT FROM AUTHOR]

Accurate identification of diseases and insect pests has been the key link to the yield, quality and safety of crops in modern agriculture. However, the same category of diseases and insect pests showed obvious differences in intra-class representation and slight similarities in inter-class representation of various diseases, due to the influence of environmental conditions, disease cycle and damaged tissues. At present, the traditional deep transfer learning methods have been difficult to cope with large-scale, multi-class fine-grained identification of pests and diseases, particularly unsuitable to the practice in complicated scenes. This paper aims to apply multiple source cameras in agricultural Internet of Things (IoT) and different intelligent precision equipment, including picking robots and smart phones, to capture high resolution 122,000 images of insect pests and diseases, covering a total of 181 fine-grained categories, including 49 types pests and 77 types diseases on different plant parts of different crops. A fine-grained recognition model was then proposed for pests and diseases based on the Multi-Stream Gaussian Probability Fusion Network (MPFN). In detail, a data-augmented method was first employed to enlarge the dataset, and then to pretrain the basic VGG19 and ResNet networks on high-quality images, in order to learn common and domain knowledge, as well fine-tuning with professional skill. Next, the refined multiple deep learning networks, including Fast-MPN and NTS-Net with transfer learning, were applied to design a multi-stream feature extractor, utilizing the mixture-granularity information to exploit high-dimensionality features, thereby to distinguish interclass discrepancy and tolerate intra-class variances. Finally, an integrated optimization was developed combining the NetVLAD feature aggregation layer with the gaussian probability fusion layer, in order to fuse various components model with Gaussian distribution as a unified probability representation for the ultimate fine-grained recognition. The input of this module was the various features of multi-models, whereas, the output was the fused classification probability. The end-to-end implementation of framework included an inner loop about the expectation maximization algorithm within an outer loop with the gradient back-propagation optimization of the whole network, indicating multi-model fusion information complementation and confidence for the overall model. The experimental results demonstrated that the MPFN model presented the excellent performance in the average recognition accuracy rate of 93.18% for a total of 181 classes of pests and diseases, indicating 5.6 percentage points better than that of the coarse-grained and fine-grained deep learning methods. In terms of test time, the average processing time of the MPFN was 61ms, indicating the basic needs of fine-grained image recognition of pests and diseases at the terminals of the IoT and intelligent equipment. In the contradistinctive analysis, training loss curves and various sub-categories identification showed that the feature aggregation fusion and Gaussian probability fusion can greatly enhance the efficiency of model with the training speed, recognition accuracy and generalized robustness for fine-grained identification of crop pests and diseases in practical scenarios. Therefore, this work can provide a technical application reference for the intelligent recognition of pests and diseases in agricultural production. In the follow-up study, more images will be taken under natural conditions to develop a more robust MPFN model deployed on actual IoTs or equipment for the pre-warning, prevention, and control of crop pests and diseases in modern agriculture. [ABSTRACT FROM AUTHOR]