Search Results (13)

A unified theory of emotion and motivation is updated in which motivational states are states in which instrumental goal-directed actions are performed to obtain anticipated rewards or avoid punishers, and emotional states are states that are elicited when the (conditioned or unconditioned) instrumental reward or punisher is or is not received. This advances our understanding of emotion and motivation, for the same set of genes and associated brain systems can define the primary or unlearned rewards and punishers such as a sweet taste or pain, and the brain systems that learn to expect rewards or punishers and that therefore produce motivational and emotional states. It is argued that instrumental actions under the control of the goal are important for emotion, because they require an intervening emotional state in which an action is learned or performed to obtain the goal, that is, the reward, or to avoid the punisher. The primate including human orbitofrontal cortex computes the reward value, and the anterior cingulate cortex is involved in learning the action to obtain the goal. In contrast, when the instrumental response is overlearned and becomes a habit with stimulus–response associations, emotional states may be less involved. In another route to output, the human orbitofrontal cortex has effective connectivity to the inferior frontal gyrus regions involved in language and provides a route for declarative reports about subjective emotional states to be produced. Reasoning brain systems provide alternative strategies to obtain rewards or avoid punishers and can provide different goals for action compared to emotional systems. Full article

This paper proposes an improved multi-agent deep deterministic policy gradient algorithm called the equal-reward and action-enhanced multi-agent deep deterministic policy gradient (EA-MADDPG) algorithm to solve the guidance problem of multiple missiles cooperating to intercept a single intruding UAV in three-dimensional space. The key innovations of EA-MADDPG include the implementation of the action filter with additional reward functions, optimal replay buffer, and equal reward setting. The additional reward functions and the action filter are set to enhance the exploration performance of the missiles during training. The optimal replay buffer and the equal reward setting are implemented to improve the utilization efficiency of exploration experiences obtained through the action filter. In order to prevent over-learning from certain experiences, a special storage mechanism is established, where experiences obtained through the action filter are stored only in the optimal replay buffer, while normal experiences are stored in both the optimal replay buffer and normal replay buffer. Meanwhile, we gradually reduce the selection probability of the action filter and the sampling ratio of the optimal replay buffer. Finally, comparative experiments show that the algorithm enhances the agents’ exploration capabilities, allowing them to learn policies more quickly and stably, which enables multiple missiles to complete the interception task more rapidly and with a higher success rate. Full article

During the vegetation growing season, the forest in the remote sensing image is more distinguishable from other background features, and the forest features are obvious and can show prominent forest area characteristics. However, deep convolutional neural network-based methods tend to overlearn the forest features in the forest extraction task, which leads to the extraction speed still having a large amount of room for improvement. In this paper, a convolutional neural network-based model is proposed based on the incorporation of spatial and channel reconstruction convolution in the U-Net model for forest extraction from remote sensing images. The network obtained an extraction accuracy of 81.781% in intersection over union (IoU), 91.317% in precision, 92.177% in recall, and 91.745% in F1-score, with a maximum improvement of 0.442% in precision when compared with the classical U-Net network. In addition, the speed of the model’s forest extraction has been improved by about 6.14 times. On this basis, we constructed a forest land dataset with high-intraclass diversity and fine-grained scale by selecting some Sentinel-2 images in Northeast China. The spatial and temporal evolutionary changes of the forest cover in the Fuxin region of Liaoning province, China, from 2019 to 2023, were obtained using this region as the study area. In addition, we obtained the change of the forest landscape pattern evolution in the Fuxin region from 2019 to 2023 based on the morphological spatial pattern analysis (MSPA) method. The results show that the core area of the forest landscape in the Fuxin region has shown an increasing change, and the non-core area has been decreasing. The SC-UNet method proposed in this paper can realize the high-precision and rapid extraction of forest in a wide area, and at the same time, it can provide a basis for evaluating the effectiveness of ecosystem restoration projects. Full article

Deep learning techniques for medical image analysis often encounter domain shifts between source and target data. Most existing approaches focus on unsupervised domain adaptation (UDA). However, in practical applications, many source domain data are often inaccessible due to issues such as privacy concerns. For instance, data from different hospitals exhibit domain shifts due to equipment discrepancies, and data from both domains cannot be accessed simultaneously because of privacy issues. This challenge, known as source-free UDA, limits the effectiveness of previous UDA medical methods. Despite the introduction of various medical source-free unsupervised domain adaptation (MSFUDA) methods, they tend to suffer from an over-fitting problem described as “longer training, worse performance”. To address this issue, we proposed the Stable Learning (SL) strategy. SL is a method that can be integrated with other approaches and consists of weight consolidation and entropy increase. Weight consolidation helps retain domain-invariant knowledge, while entropy increase prevents over-learning. We validated our strategy through experiments on three MSFUDA methods and two public datasets. For the abdominal dataset, the application of the SL strategy enables the MSFUDA method to effectively address the domain shift issue. This results in an improvement in the Dice coefficient from 0.5167 to 0.7006 for the adaptation from CT to MRI, and from 0.6474 to 0.7188 for the adaptation from MRI to CT. The same improvement is observed with the cardiac dataset. Additionally, we conducted ablation studies on the two involved modules, and the results demonstrated the effectiveness of the SL strategy. Full article

With the development of ultra-long-range visual sensors, the application of unsupervised person re-identification algorithms to them has become increasingly important. However, these algorithms inevitably generate noisy pseudo-labels, which seriously hinder the performance of tasks over a large range. Mixup, a data enhancement technique, has been validated in supervised learning for its generalization to noisy labels. Based on this observation, to our knowledge, this study is the first to explore the impact of the mixup technique on unsupervised person re-identification, which is a downstream task of contrastive learning, in detail. Specifically, mixup was applied in different locations (at the pixel level and feature level) in an unsupervised person re-identification framework to explore its influences on task performance. In addition, based on the richness of the information contained in the person samples to be mixed, we propose an uncertainty-aware mixup (UnA-Mix) method, which reduces the over-learning of simple person samples and avoids the information damage that occurs when information-rich person samples are mixed. The experimental results on three benchmark person re-identification datasets demonstrated the applicability of the proposed method, especially on the MSMT17, where it outperformed state-of-the-art methods by 5.2% and 4.8% in terms of the mAP and rank-1, respectively. Full article

Continuous data stream analysis primarily focuses on the unanticipated changes in the transmission of data distribution over time. Conceptual change is defined as the signal distribution changes over the transmission of continuous data streams. A drift detection scenario is set forth to develop methods and strategies for detecting, interpreting, and adapting to conceptual changes over data streams. Machine learning approaches can produce poor learning outcomes in the conceptual change environment if the sudden change is not addressed. Furthermore, due to developments in concept drift, learning methodologies have been significantly systematic in recent years. The research introduces a novel approach using the fully connected committee machine (FCM) and different activation functions to address conceptual changes in continuous data streams. It explores scenarios of continual learning and investigates the effects of over-learning and weight decay on concept drift. The findings demonstrate the effectiveness of the FCM framework and provide insights into improving machine learning approaches for continuous data stream analysis. We used a layered neural network framework to experiment with different scenarios of continual learning on continuous data streams in the presence of change in the data distribution using a fully connected committee machine (FCM). In this research, we conduct experiments in various scenarios using a layered neural network framework, specifically the fully connected committee machine (FCM), to address conceptual changes in continuous data streams for continual learning under a conceptual change in the data distribution. Sigmoidal and ReLU (Rectified Linear Unit) activation functions are considered for learning regression in layered neural networks. When the layered framework is trained from the input data stream, the regression scheme changes consciously in all scenarios. A fully connected committee machine (FCM) is trained to perform the tasks described in continual learning with M hidden units on dynamically generated inputs. In this method, we run Monte Carlo simulations with the same number of units on both sides, K and M, to define the advancement of intersections between several hidden units and the calculation of generalization error. This is applied to over-learnability as a method of over-forgetting, integrating weight decay, and examining its effects when a concept drift is presented. Full article

The temporal subtraction technique is a useful tool for computer aided diagnosis (CAD) in visual screening. The technique subtracts the previous image set from the current one for the same subject to emphasize temporal changes and/or new abnormalities. However, it is difficult to obtain a clear subtraction image without subtraction image artifacts. VoxelMorph in deep learning is a useful method, as preparing large training datasets is difficult for medical image analysis, and the possibilities of incorrect learning, gradient loss, and overlearning are concerns. To overcome this problem, we propose a new method for generating temporal subtraction images of thoracic multi-detector row computed tomography (MDCT) images based on ResidualVoxelMorph, which introduces a residual block to VoxelMorph to enable flexible positioning at a low computational cost. Its high learning efficiency can be expected even with a limited training set by introducing residual blocks to VoxelMorph. We performed our method on 84 clinical images and evaluated based on three-fold cross-validation. The results showed that the proposed method reduced subtraction image artifacts on root mean square error (RMSE) by 11.3% (p < 0.01), and its effectiveness was verified. That is, the proposed temporal subtraction method for thoracic MDCT improves the observer’s performance. Full article

Semantic segmentation for high-resolution remote-sensing imagery (HRRSI) has become increasingly popular in machine vision in recent years. Most of the state-of-the-art methods for semantic segmentation of HRRSI usually emphasize the strong learning ability of deep convolutional neural network to model the contextual relationship in the image, which takes too much consideration on every pixel in images and subsequently causes the problem of overlearning. Annotation errors and easily confused features can also lead to the confusion problem while using the pixel-based methods. Therefore, we propose a new semantic segmentation network—the region-enhancing network (RE-Net)—to emphasize the regional information instead of pixels to solve the above problems. RE-Net introduces the regional information into the base network, to enhance the regional integrity of images and thus reduce misclassification. Specifically, the regional context learning procedure (RCLP) can learn the context relationship from the perspective of regions. The region correcting procedure (RCP) uses the pixel aggregation feature to recalibrate the pixel features in each region. In addition, another simple intra-network multi-scale attention module is introduced to select features at different scales by the size of the region. A large number of comparative experiments on four different public datasets demonstrate that the proposed RE-Net performs better than most of the state-of-the-art ones. Full article

In reading, binocular eye movements are required for optimal visual processing and thus, in case of asthenopia or reading problems, standard orthoptic and optometric routines check individual binocular vision by a variety of tests. The present study therefore examines the predictive value of such standard measures of heterophoria, accommodative and vergence facility, AC/A-ratio, NPC and symptoms for binocular coordination parameters during reading. Binocular eye movements were recorded (EyeLink II) for 65 volunteers during a typical reading task and linear regression analyses related all parameters of binocular coordination to all above-mentioned optometric measures: while saccade disconjugacy was weakly predicted by vergence facility (15% explained variance), vergence facility, AC/A and symptoms scores predicted vergence drift (31%). Heterophoria, vergence facility and NPC explained 31% of fixation disparity and first fixation duration showed minor relations to symptoms (18%). In sum, we found only weak to moderate relationships, with expected, selective associations: dynamic parameter related to optometric tests addressing vergence dynamics, whereas the static parameter (fixation disparity) related mainly to heterophoria. Most surprisingly, symptoms were only loosely related to vergence drift and fixation duration, reflecting associations to a dynamic aspect of binocular eye movements in reading and potentially non-specific, overall but slight reading deficiency. Thus, the efficiency of optometric tests to predict binocular coordination during reading was low—questioning a simple, straightforward extrapolation of such test results to an overlearned, complex task. Full article

Numerous studies have noted that sex and/or menstrual phase influences cognitive performance (in particular, declarative memory), but the effects on motor learning (ML) and procedural memory/consolidation remain unclear. In order to test the hypothesis that ML differs across menstrual cycle phases, initial ML, overlearning, consolidation, and final performance were assessed in women in the follicular, preovulation and luteal phases. Primary motor cortex (M1) oscillations were assessed neuro-physiologically, and premenstrual syndrome and interoceptive awareness scores were assessed psychologically. We found not only poorer performance gain through initial ML but also lower final performance after overlearning a day and a week later in the luteal group than in the ovulation group. This behavioral difference could be explained by particular premenstrual syndrome symptoms and associated failure of normal M1 excitability in the luteal group. In contrast, the offline effects, i.e., early and late consolidation, did not differ across menstrual cycle phases. These results provide information regarding the best time in which to start learning new sensorimotor skills to achieve expected gains. Full article

Attention is classically classified according to mode of engagement into voluntary and reflexive, and type of operation into covert and overt. The first distinguishes whether attention is elicited intentionally or by unexpected events; the second, whether attention is directed with or without eye movements. Recently, this taxonomy has been expanded to include automated orienting engaged by overlearned symbols and combined attention engaged by a combination of several modes of function. However, so far, combined effects were demonstrated in covert conditions only, and, thus, here we examined if attentional modes combined in overt responses as well. To do so, we elicited automated, voluntary, and combined orienting in covert, i.e., when participants responded manually and maintained central fixation, and overt cases, i.e., when they responded by looking. The data indicated typical effects for automated and voluntary conditions in both covert and overt data, with the magnitudes of the combined effect larger than the magnitude of each mode alone as well as their additive sum. No differences in the combined effects emerged across covert and overt conditions. As such, these results show that attentional systems combine similarly in covert and overt responses and highlight attention’s dynamic flexibility in facilitating human behavior. Full article

Artificial neural networks are widely applied for prediction, function simulation, and data classification. Among these applications, the wavelet neural network is widely used in image classification problems due to its advantages of high approximation capabilities, fault-tolerant capabilities, learning capacity, its ability to effectively overcome local minimization issues, and so on. The error function of a network is critical to determine the convergence, stability, and classification accuracy of a neural network. The selection of the error function directly determines the network’s performance. Different error functions will correspond with different minimum error values in training samples. With the decrease of network errors, the accuracy of the image classification is increased. However, if the image classification accuracy is difficult to improve upon, or is even decreased with the decreasing of the errors, then this indicates that the network has an “over-learning” phenomenon, which is closely related to the selection of the function errors. With regards to remote sensing data, it has not yet been reported whether there have been studies conducted regarding the “over-learning” phenomenon, as well as the relationship between the “over-learning” phenomenon and error functions. This study takes SAR, hyper-spectral, high-resolution, and multi-spectral images as data sources, in order to comprehensively and systematically analyze the possibility of an “over-learning” phenomenon in the remote sensing images from the aspects of image characteristics and neural network. Then, this study discusses the impact of three typical entropy error functions (NB, CE, and SH) on the “over-learning” phenomenon of a network. The experimental results show that the “over-learning” phenomenon may be caused only when there is a strong separability between the ground features, a low image complexity, a small image size, and a large number of hidden nodes. The SH entropy error function in that case will show a good “over-learning” resistance ability. However, for remote sensing image classification, the “over-learning” phenomenon will not be easily caused in most cases, due to the complexity of the image itself, and the diversity of the ground features. In that case, the NB and CE entropy error network mainly show a good stability. Therefore, a blind selection of a SH entropy error function with a high “over-learning” resistance ability from the wavelet neural network classification of the remote sensing image will only decrease the classification accuracy of the remote sensing image. It is therefore recommended to use an NB or CE entropy error function with a stable learning effect. Full article

Findings from recordings of human temporal cortical single neuron activity during several measures of language, including object naming and word reading are reviewed and related to changes in activity in the same neurons during recent verbal memory and verbal associative learning measures, in studies conducted during awake neurosurgery for the treatment of epilepsy. The proportion of neurons changing activity with language tasks was similar in either hemisphere. Dominant hemisphere activity was characterized by relative inhibition, some of which occurred during overt speech, possibly to block perception of one’s own voice. However, the majority seems to represent a dynamic network becoming active with verbal memory encoding and especially verbal learning, but inhibited during performance of overlearned language tasks. Individual neurons are involved in different networks for different aspects of language, including naming or reading and naming in different languages. The majority of the changes in activity were tonic sustained shifts in firing. Patterned phasic activity for specific language items was very infrequently recorded. Human single neuron recordings provide a unique perspective on the biologic substrate for language, for these findings are in contrast to many of the findings from other techniques for investigating this. Full article