RESUMEN

Spiking neural networks (SNNs) have received increasing attention due to their high biological plausibility and energy efficiency. The binary spike-based information propagation enables efficient sparse computation in event-based and static computer vision applications. However, the weight precision and especially the membrane potential precision remain as high-precision values (e.g., 32 bits) in state-of-the-art SNN algorithms. Each neuron in an SNN stores the membrane potential over time and typically updates its value in every time step. Such frequent read/write operations of high-precision membrane potential incur storage and memory access overhead in SNNs, which undermines the SNNs' compatibility with resource-constrained hardware. To resolve this inefficiency, prior works have explored the time step reduction and low-precision representation of membrane potential at a limited scale and reported significant accuracy drops. Furthermore, while recent advances in on-device AI present pruning and quantization optimization with different architectures and datasets, simultaneous pruning with quantization is highly under-explored in SNNs. In this work, we present SpQuant-SNN, a fully-quantized spiking neural network with ultra-low precision weights, membrane potential, and high spatial-channel sparsity, enabling the end-to-end low precision with significantly reduced operations on SNN. First, we propose an integer-only quantization scheme for the membrane potential with a stacked surrogate gradient function, a simple-yet-effective method that enables the smooth learning process of quantized SNN training. Second, we implement spatial-channel pruning with membrane potential prior, toward reducing the layer-wise computational complexity, and floating-point operations (FLOPs) in SNNs. Finally, to further improve the accuracy of low-precision and sparse SNN, we propose a self-adaptive learnable potential threshold for SNN training. Equipped with high biological adaptiveness, minimal computations, and memory utilization, SpQuant-SNN achieves state-of-the-art performance across multiple SNN models for both event-based and static image datasets, including both image classification and object detection tasks. The proposed SpQuant-SNN achieved up to 13× memory reduction and >4.7× FLOPs reduction with < 1.8% accuracy degradation for both classification and object detection tasks, compared to the SOTA baseline.

RESUMEN

OBJECTIVES: For constructing an isolated tooth identification system using deep learning, Igarashi et al. (2021) began constructing a learning model as basic research to identify the left and right mandibular first and second premolars. These teeth were chosen for analysis because they are difficult to identify from one another. The learning method itself was proven appropriate but presented low accuracy. Therefore, further improvement in the learning data should increase the accuracy of the model. The study objectives were to modify the learning data and increase the learning model accuracy for enabling the identification of isolated lower premolars. METHODS: Static images of the occlusal surface of the premolars made from the dental plaster casts of dental students were used as the training, validation, and test data. A convolutional neural network with 32 hidden layers, AlexNet, convolutional architecture for fast feature embedding, and stochastic gradient descent was used to construct four learning models. RESULTS: The accuracy of the identification model increased using static images of the occlusal surface of the teeth with the adjacent teeth deleted as the training and validation data; however, a learning model that could perfectly identify the teeth could not be realized. CONCLUSIONS: Static images of the occlusal surface of the teeth with the adjacent teeth deleted should be used as both training and validation data. The ratio of the numbers of training, validation, and test data should be optimized.

Asunto(s)

RESUMEN

Internal bodily signals provide an essential function for human survival. Accurate recognition of such signals in the self, known as interoception, supports the maintenance of homeostasis, and is closely related to emotional processing, learning and decision-making, and mental health. While numerous studies have investigated interoception in the self, the recognition of these states in others has not been examined despite its crucial importance for successful social relationships. This paper presents the development and validation of the Interoceptive States Static Images (ISSI), introducing a validated database of 423 visual stimuli for the study of non-affective internal state recognition in others, freely available to other researchers. Actors were photographed expressing various exemplars of both interoceptive states and control actions. The images went through a two-stage validation procedure, the first involving free-labelling and the second using multiple choice labelling and quality rating scales. Five scores were calculated for each stimulus, providing information about the quality and specificity of the depiction, as well as the extent to which labels matched the intended state/action. Results demonstrated that control action stimuli were more recognisable than internal state stimuli. Inter-category variability was found for the internal states, with some states being more recognisable than others. Recommendations for the utilisation of ISSI stimuli are discussed. The stimulus set is freely available to researchers, alongside data concerning recognisability.

Asunto(s)

RESUMEN

BACKGROUND: Successful identification of emotional expression in patients is of considerable importance in the diagnosis of diseases and while developing rapport between physicians and patients. Despite the importance of such skills, this aspect remains grossly overlooked in conventional medical training in India. This study aims to explore the extent to which medical students can identify emotions by observing photographs of male and female subjects expressing different facial expressions. METHODS: A total of 106 medical students aged 18-25, without any diagnosed mental illnesses, were shown images of the six universal facial expressions (anger, sadness, fear, happiness, disgust, and surprise) at 100% intensity with an exposure time of 2 seconds for each image. The participants marked their responses after each image was shown. Collected data were analyzed using Statistical Package for the Social Sciences. RESULTS: Participants could identify 76.54% of the emotions on average, with higher accuracy for positive emotions (95.6% for happiness) and lower for negative emotions (46% for fear). There were no significant variations in identification with respect to sex of the observers. However, it was seen that participants could identify emotions better from male faces than those from female faces, a finding that was statistically significant. Negative emotions were identified more accurately from male faces, while positive emotions were identified better from female ones. CONCLUSIONS: Male participants identified emotions better from male faces, while females identified positive emotions better from female faces and negative ones from male faces.

RESUMEN

Time perception can be affected by real emotional pictures of people that evoke different levels of arousal. Figurative artwork images of body postures that imply movement with different intensities and evoke different levels of arousal can modulate the perception of time. The present study investigated whether abstract paintings that represent motion in different ways affect the perception of time when subjects are exposed to the paintings for different durations. Undergraduate students observed 20 abstract paintings from different artistic schools (i.e., cubism, constructivism, expressionism, and futurism). They observed for 3 s each painting and estimated the time of exposure (reproduction method). After the time estimations, the subjects completed different semantic Movement, Arousal, Complexity, and Recognition scales to obtain information about how the painting compositions were perceived. Time distortions were observed for only two cubist paintings that represented human forms, which were related to both evoked arousal and implied movement (Experiment 1). Experiment 2 further verified whether these time distortions were related to implied movement perception or arousal. Different groups of participants were exposed for 3 and 9 s to only four cubist paintings that represented human forms. These time exposures (3 and 9 s) were used because the arousal-evoking effects may be transient for exposure times that are longer than 2-3 s. The data analysis revealed overestimation of time for the cubist painting that had greater arousal and movement scores only when the subjects were exposed for 9 s, showing that implied movement in abstract human figures is more effective than images with emotional content. We discuss the effect of durations of exposure to pictorial characteristics of artwork on time perception, emphasizing aspects of the visual perception of human forms in cubist paintings and arousal effects in an aesthetic episode.(AU)

Asunto(s)

RESUMEN

Time perception can be affected by real emotional pictures of people that evoke different levels of arousal. Figurative artwork images of body postures that imply movement with different intensities and evoke different levels of arousal can modulate the perception of time. The present study investigated whether abstract paintings that represent motion in different ways affect the perception of time when subjects are exposed to the paintings for different durations. Undergraduate students observed 20 abstract paintings from different artistic schools (i.e., cubism, constructivism, expressionism, and futurism). They observed for 3 s each painting and estimated the time of exposure (reproduction method). After the time estimations, the subjects completed different semantic Movement, Arousal, Complexity, and Recognition scales to obtain information about how the painting compositions were perceived. Time distortions were observed for only two cubist paintings that represented human forms, which were related to both evoked arousal and implied movement (Experiment 1). Experiment 2 further verified whether these time distortions were related to implied movement perception or arousal. Different groups of participants were exposed for 3 and 9 s to only four cubist paintings that represented human forms. These time exposures (3 and 9 s) were used because the arousal-evoking effects may be transient for exposure times that are longer than 2-3 s. The data analysis revealed overestimation of time for the cubist painting that had greater arousal and movement scores only when the subjects were exposed for 9 s, showing that implied movement in abstract human figures is more effective than images with emotional content. We discuss the effect of durations of exposure to pictorial characteristics of artwork on time perception, emphasizing aspects of the visual perception of human forms in cubist paintings and arousal effects in an aesthetic episode.

Asunto(s)

RESUMEN

OBJECTIVE: To describe and validate the simulation of the basic features of GE Millennium MG gamma camera using the GATE Monte Carlo platform. MATERIAL AND METHODS: Crystal size and thickness, parallel-hole collimation and a realistic energy acquisition window were simulated in the GATE platform. GATE results were compared to experimental data in the following imaging conditions: a point source of (99m)Tc at different positions during static imaging and tomographic acquisitions using two different energy windows. The accuracy between the events expected and detected by simulation was obtained with the Mann-Whitney-Wilcoxon test. Comparisons were made regarding the measurement of sensitivity and spatial resolution, static and tomographic. Simulated and experimental spatial resolutions for tomographic data were compared with the Kruskal-Wallis test to assess simulation accuracy for this parameter. RESULTS: There was good agreement between simulated and experimental data. The number of decays expected when compared with the number of decays registered, showed small deviation (≤ 0.007%). The sensitivity comparisons between static acquisitions for different distances from source to collimator (1, 5, 10, 20, 30 cm) with energy windows of 126-154 keV and 130-158 keV showed differences of 4.4%, 5.5%, 4.2%, 5.5%, 4.5% and 5.4%, 6.3%, 6.3%, 5.8%, 5.3%, respectively. For the tomographic acquisitions, the mean differences were 7.5% and 9.8% for the energy window 126-154 keV and 130-158 keV. Comparison of simulated and experimental spatial resolutions for tomographic data showed no statistically significant differences with 95% confidence interval. CONCLUSIONS: Adequate simulation of the system basic features using GATE Monte Carlo simulation platform was achieved and validated.

Asunto(s)

RESUMEN

When a video of someone speaking is paused, the stationary image of the speaker typically appears less flattering than the video, which contained motion. We call this the frozen face effect (FFE). Here we report six experiments intended to quantify this effect and determine its cause. In Experiment 1, video clips of people speaking in naturalistic settings as well as all of the static frames that composed each video were presented, and subjects rated how flattering each stimulus was. The videos were rated to be significantly more flattering than the static images, confirming the FFE. In Experiment 2, videos and static images were inverted, and the videos were again rated as more flattering than the static images. In Experiment 3, a discrimination task measured recognition of the static images that composed each video. Recognition did not correlate with flattery ratings, suggesting that the FFE is not due to better memory for particularly distinct images. In Experiment 4, flattery ratings for groups of static images were compared with those for videos and static images. Ratings for the video stimuli were higher than those for either the group or individual static stimuli, suggesting that the amount of information available is not what produces the FFE. In Experiment 5, videos were presented under four conditions: forward motion, inverted forward motion, reversed motion, and scrambled frame sequence. Flattery ratings for the scrambled videos were significantly lower than those for the other three conditions. In Experiment 6, as in Experiment 2, inverted videos and static images were compared with upright ones, and the response measure was changed to perceived attractiveness. Videos were rated as more attractive than the static images for both upright and inverted stimuli. Overall, the results suggest that the FFE requires continuous, natural motion of faces, is not sensitive to inversion, and is not due to a memory effect.