Temporal Position Research Articles

Multi-step prediction of greenhouse temperature and humidity based on temporal position attention LSTM

Multi-step prediction of greenhouse temperature and humidity based on temporal position attention LSTM

Reinforcement Learning with Multi-Policy Movement Strategy for Weakly Supervised Temporal Sentence Grounding

Temporal grounding involves identifying the target moment based on the provided sentence in an untrimmed video. In weakly supervised temporal grounding studies, existing temporal sentence grounding methods face challenges in (1) learning semantic alignment between the candidate window and language query and (2) identifying accurate temporal boundaries during the grounding process. In this work, we propose a reinforcement learning (RL)-based multi-policy movement framework (MMF) for weakly supervised temporal sentence grounding. We imitate the behavior of human beings when grounding specified content in a video, starting from a coarse location and then identifying fine-grained temporal boundaries. The RL-based framework initially sets a series of candidate windows and learns to adjust them step-by-step by maximizing the rewards, indicating the semantic alignment between the current window and the query. To better learn the alignment, we propose a Gaussian-based Dual-Alignment Module (GDAM) which combines the strengths of both scoring-based and reconstruction-based alignment methods, addressing the issues of negative sample bias and language bias. We also employ the multi-policy movement strategy (MMS) which grounds the temporal position in a coarse-to-fine manner. Extensive experiments demonstrate that our proposed method outperforms existing weakly supervised algorithms, achieving state-of-the-art performance on the Charades-STA and ActivityNet Captions datasets.

Dynamic Attending in percussionists and wind players

Percussionists have been shown to have superior temporal perception, but the underlying mechanisms for this advantage deserve further investigation. Dynamic Attending Theory, which posits the way attention aligns with rhythmic temporal structure, may explain why percussionists exhibit superior temporal perception. Percussionists may entrain more strongly to periodic stimuli than other musicians, and possibly at multiple hierarchical levels as opposed to only the tactus. To test these hypotheses, we asked a total of 41 musicians, 20 percussionists, and 21 wind players to complete a target detection task. Participants listened to periodic bursts of noise ( priming bursts) and tried to detect target bursts of noise at different temporal positions, some in phase and some out of phase with the initial noises. Target-burst amplitudes were calibrated to be at each participant’s perceptual threshold. We analyzed the hit rates at the different temporal positions, as well as the Fourier transform of those hit rate curves. No discernible difference between groups was found in the time domain, but in the frequency domain the percussionist group showed stronger fluctuation at a frequency that matched the priming bursts, but not at other hierarchical levels (specifically, double that frequency). Our findings help explain percussionists’ perceptual advantages, although questions about entrainment at different hierarchical levels deserve further consideration.

Rhythm-based Temporal Expectations: Unique Contributions of Predictability and Periodicity.

Anticipating events and focusing attention accordingly are crucial for navigating our dynamic environment. Rhythmic patterns of sensory input offer valuable cues for temporal expectations and facilitate perceptual processing. Rhythm-based temporal expectations may rely on oscillatory entrainment, where neural activity and perceptual sensitivity synchronize with periodic stimuli. However, whether entrainment models can account for aperiodic predictable rhythms remains unclear. Our study aimed to delineate the distinct roles of predictability and periodicity in rhythm-based expectations. Participants performed a pitch-identification task preceded by periodic predictable, aperiodic predictable, or aperiodic unpredictable temporal sequences. By manipulating the temporal position of the target sound, we observed how auditory perceptual performance was modulated by the target position's relative phase relationship to the preceding sequences. Results revealed a significant performance advantage for predictable sequences, both periodic and aperiodic, compared with unpredictable ones. However, only the periodic sequence induced an entrained modulation pattern, with performance peaking in synchrony with the inherent sequence continuation. Event-related brain potentials corroborated these findings. The target-evoked P3b, possibly a neural marker of attention allocation, mirrored the behavioral performance patterns. This supports our hypothesis that temporal attention guided by rhythm-based expectations modulates perceptual performance. Furthermore, the predictive sequences were associated with enhanced target-preceding negativity (akin to the contingent negative variation), indicating enhanced target preparation. The periodic-specific modulation likely reflects more precise temporal expectations, potentially involving neural entrainment and/or more focused attention. Our findings suggest that predictability and periodicity influence perception through distinct mechanisms.

How It Begins and Ends: Temporal Positions of Linguistic Indicators in Staged and Authentic Emergency Calls

AbstractThe utilisation of emergency calls to stage false and misleading narratives can assist offenders in concealing their criminal activity. The majority of emergency call analyses focus on detecting deception; however, investigating the complex temporal ordering of verbal indicators with a measure of proximity may offer new inductive investigative insights. The current research took a novel approach to analysing emergency calls by focusing on the temporal placement of verbal indicators with a measure of proximity coefficients. Results showed that several important differences existed in the temporal ordering of indicators that occur at the start and end of staged and authentic calls. These findings offer an insight as to influence of saliency and the competing priorities that exist between staged and authentic emergency calls. This is the first research to explore the temporal ordering of indicators in emergency calls and to provide a basis for its tentative application to interview strategies.

A preserved neural code for temporal order between memory formation and recall in the human medial temporal lobe.

Temporal memory enables us to remember the temporal order of events happening in our life. The human medial temporal lobe (MTL) appears to contain neural representations supporting temporal memory formation, but the cellular mechanisms that preserve temporal order information for recall are largely unknown. Here, we examined whether human MTL neuronal activity represents the temporal position of events during memory formation and recall, using invasive single and multi-unit recordings in human epilepsy patients (n = 19). Participants freely navigated a virtual environment in order to explore and remember locations and temporal positions of objects. During each exploration period, they sequentially encountered two or three different objects, placed in different locations. This allowed us to examine single- and multi-unit neuronal firing rates (FR) as a function of the temporal position the objects were presented in. We found that a significant number of multi-units and single-units in various MTL regions including the hippocampus showed selectivity to the temporal position of objects during the exploration period. During recall, patients were asked to indicate which one of two objects from the same trial was found latter. Neural firing rates during recall showed a selectivity supporting recall of temporal positions. Interestingly, most of the selective single-units that stayed selective during encoding and recall preserved their temporal position preference. Our results thus suggest that neuronal activity in the human MTL contains a preserved neural code for temporal order in memory formation and recall.

Distinct effects of slow and fast theta tACS in enhancing temporal memory

Abstract Temporal memory plays a crucial role in organizing real-life events into meaningful segments. Previous research suggests that the clustering of temporally related information builds on the neural correlates of associative memory, including theta oscillations. Transcranial alternating current stimulation (tACS) provides a means of modulating theta oscillations within associative memory networks, possibly including hippocampal modulation when targeting parietal cortex. Theta oscillations are not limited to a single frequency range, but instead, exhibit functional specialization, with slow theta (3 Hz) implicated in short-term episodic memory formation and fast theta (8 Hz) involved in spatial navigation. This study aimed to investigate the distinct effects of slow and fast theta stimulation on temporal memory. Participants encoded visual objects paired with frame color while receiving tACS stimulation at 8 Hz, 3 Hz, or sham targeting the left parietal cortex. The frame color would change for every eight objects, establishing a context boundary with each color change. Subsequently, participants performed a timeline task to assess temporal memory performance. Results showed that slow, but not fast, theta stimulation led to an enhancement in temporal accuracy (absolute temporal error) compared to sham tACS, in support of our main hypothesis. Under sham stimulation, participants consistently underestimated the temporal position of items presented further away from boundary, compared to those presented at boundary. This finding resembled temporal compression observed during event segmentation. Interestingly, fast, but not slow, theta stimulation reduced this temporal bias (rated position— actual position). This study represents the first tACS evidence for differential contributions of slow versus fast theta to temporal memory formation in humans. We speculate that parietal theta tACS may modulate the hippocampus and facilitate temporal memory formation.

Pattern recognition using spiking antiferromagnetic neurons

Spintronic devices offer a promising avenue for the development of nanoscale, energy-efficient artificial neurons for neuromorphic computing. It has previously been shown that with antiferromagnetic (AFM) oscillators, ultra-fast spiking artificial neurons can be made that mimic many unique features of biological neurons. In this work, we train an artificial neural network of AFM neurons to perform pattern recognition. A simple machine learning algorithm called spike pattern association neuron (SPAN), which relies on the temporal position of neuron spikes, is used during training. In under a microsecond of physical time, the AFM neural network is trained to recognize symbols composed from a grid by producing a spike within a specified time window. We further achieve multi-symbol recognition with the addition of an output layer to suppress undesirable spikes. Through the utilization of AFM neurons and the SPAN algorithm, we create a neural network capable of high-accuracy recognition with overall power consumption on the order of picojoules.

Temporal latencies and position uncertainty in stereoscopic and luminance motion using a continuous eye-tracking task

Temporal latencies and position uncertainty in stereoscopic and luminance motion using a continuous eye-tracking task

A Multi-Scale Temporal Feature Extraction Approach for Network Traffic Anomaly Detection

With the rapid advancement of network technology and the expanding scale of the Internet, network traffic data has grown exponentially, leading to increasingly prominent issues in network security. Many researchers are developing new network traffic anomaly detection models, but these models are difficult to effectively capture the multi-scale temporal characteristics of network traffic data and learn the correlation and importance among various feature dimensions. To this end, we propose a multi-scale temporal feature network (MSTFN-AM) based on an attention mechanism. MSTFN-AM integrates original data with temporal information through the temporal position encoding module and extracts multi-scale temporal features through the temporal feature extraction module. The temporal self-attention module can effectively identify dependencies and correlations between different time points. Ultimately, experiments conducted on two public datasets demonstrate that the MSTFN-AM model outperforms all baseline models in prediction performance.

The Hippocampus Represents Information about Movements in Their Temporal Position in a Learned Motor Sequence.

Our repertoire of motor skills is filled with sequential movements that need to be performed in a specific order. Here, we used functional magnetic resonance imaging to investigate whether the human hippocampus, a region known to support temporal order in non-motor memory, represents information about the order of sequential motor actions in human participants (both sexes). We also examined such representations in other regions of the motor network (i.e., the premotor cortex, supplementary motor area, anterior superior parietal lobule, and striatum) already known for their critical role in motor sequence learning. Results showed that the hippocampus represents information about movements in their learned temporal position in the sequence, but not about movements or temporal positions in random movement patterns. Other regions of the motor network coded for movements in their learned temporal position, as well as movements and positions in random movement patterns. Importantly, movement coding contributed to sequence learning patterns in primary, supplementary, and premotor cortices but not in striatal and parietal regions. Our findings deepen our understanding of how striatal and cortical regions contribute to motor sequence learning and point to the capacity of the hippocampus to represent movements in their temporal context, an ability possibly explaining its contribution to motor learning.

Multidimensional time series motif group discovery based on matrix profile

With the continuous advancements in sensor technology and the increasing capabilities for data collection and storage, the acquisition of time series data across diverse domains has become significantly easier. Consequently, there is a growing demand for identifying potential motifs within multidimensional time series. The introduction of the Matrix Profile (MP) structure and the mSTOMP algorithm enables the detection of multidimensional motifs in large-scale time series datasets. However, the Matrix Profile (MP) does not provide information regarding the frequency of occurrence of these motifs. As a result, it is challenging to determine whether a motif appears frequently or to identify the specific time periods during which it typically occurs, thereby limiting further analysis of the discovered motifs. To address this limitation, we proposed Index Link Motif Group Discovery (ILMGD) algorithm, which uses index linking to rapidly merge and group multidimensional motifs. Based on the results of the ILMGD algorithm, we can determine the frequency and temporal positions of motifs, facilitating deeper analysis. Our proposed method requires minimal additional parameters and reduces the need for extensive manual intervention. We validate the effectiveness of our algorithm on synthetic datasets and demonstrate its applicability on three real-world datasets, highlighting how it enables a comprehensive understanding of the discovered motifs.

Transformer models for Land Cover Classification with Satellite Image Time Series

In this paper we address the task of pixel-wise land cover (LC) classification using satellite image time series (SITS). For that purpose, we use a supervised deep learning model and focus on combining spatial and temporal features. Our method is based on the Swin Transformer and captures global temporal features by using self-attention and local spatial features by convolutions. We extend the architecture to receive multi-temporal input to generate one output label map for every input image. In our experiments we focus on the application of pixel-wise LC classification from Sentinel‑2 SITS over the whole area of Lower Saxony (Germany). The experiments with our new model show that by using convolutions for spatial feature extraction or a temporal weighting module in the skip connections the performance improves and is more stable. The combined usage of both adaptations results in the overall best performance although this improvement is only minimal. Compared to a fully convolutional neural network without any self-attention layers our model improves the results by 2.1% in the mean F1-Score on a corrected test dataset. Additionally, we investigate different types of temporal position encoding, which do not have a significant impact on the performance.

ConvTransformer Attention Network for temporal action detection

Boundary detection is a challenging problem in Temporal Action Detection (TAD). While transformer-based methods achieve satisfactory results by incorporating self-attention mechanisms to model global dependencies for boundary detection, they face two key issues. Firstly, they lack explicit learning of local relationships; this limitation results in imprecise boundary detection when subtle appearance changes occur between adjacent clips. Secondly, transformer-based methods lead to feature convergence across multiple actions due to the self-attention mechanism’s tendency to distribute focus across the entire input video, resulting in the prediction of imprecisely overlapping actions. To address these challenges, we introduce the ConvTransformer Attention Network (CTAN), a novel framework comprised of two primary components: (1) The Temporal Attention Block (TAB), a temporal attention mechanism designed to emphasize critical temporal positions enriched with essential action-related features. (2) The ConvTransformer Block (CTB), which employs a hybrid structure for capturing nuanced appearance changes locally and action transitions globally. Facilitated with these components, CTAN is adept at focusing on motion features between overlapping actions, and precisely capturing both local differences between adjacent clips and global action transitions. The extensive experiments on multiple datasets, including THUMOS14, MultiTHUMOS, and ActivityNet, confirm the effectiveness of CTAN.

Comparison of 96-kV and 120-kV cone-beam CT for the assessment of cochlear implants

BackgroundTo compare the diagnostic value of 120-kV with conventional 96-kV Cone-Beam CT (CBCT) of the temporal bone after cochlear implant (CI) surgery.MethodsThis retrospective study included CBCT scans after CI surgery between 06/17 and 01/18. CBCT allowed examinations with 96-kV or 120-kV; other parameters were the same. Two radiologists independently evaluated following criteria on 5-point Likert scales: osseous spiral lamina, inner and outer cochlear wall, semi-circular canals, mastoid trabecular structure, overall image quality, metal and motion artefacts, depiction of intracochlear electrode position and visualisation of single electrode contacts. Effective radiation dose was assessed.ResultsSeventy-five patients (females, n = 39 [52.0%], mean age, 55.8 ± 16.5 years) were scanned with 96-kV (n = 32, 42.7%) and 120-kV (n = 43, 57.3%) protocols including CI models from three vendors (vendor A n = 7; vendor B n = 43; vendor C n = 25). Overall image quality, depiction of anatomical structures, and electrode position were rated significantly better in 120-kV images compared to 96-kV (all p < = 0.018). Anatomical structures and electrode position were rated significantly better in 120-kV CBCT for CI models from vendor A and C, while 120-kV did not provide improved image quality in CI models from vendor B. Radiation doses were significantly higher for 120-kV scans compared to 96-kV (0.15 vs. 0.08 mSv, p < 0.001).Conclusions120-kV and 96-kV CBCT provide good diagnostic images for the postoperative CI evaluation. While 120-kV showed improved depiction of temporal bone and CI electrode position compared to 96-kV in most CI models, the 120-kV protocol should be chosen wisely due to a substantially higher radiation exposure.

China’s spatial and temporal position in the global research network of information behavior and its current development dynamics

PurposeThis paper aims to objectively present the research dynamics of China in the field of information behavior and its development trends. Firstly, it incorporates China’s research in the field of information behavior into the global research network of information behavior, analyzing the changes in the status of Chinese scholars and their research institutions in the global research network from 1991 to 2022, the trends in publication volume and the cooperation relationships with other countries. Then, it conducts a detailed analysis of China’s research categories, groups, theoretical models and hot topics in different information contexts in the past five years (2018–2022).Design/methodology/approachThe study retrieved research literature related to information behavior in China from 1991 to 2022 in the Web of Science database. It then utilized a national/institutional cooperation network map to analyze the changes in the status of Chinese scholars/institutions in the global research network during this period, publication volume trends and cooperation relationships with other countries. Furthermore, it employed keyword co-occurrence network maps to analyze the key categories, groups, theories and models of China’s research in different information contexts in the past five years. Based on this, it used keyword clustering network maps to analyze the hot topics of China’s research in different information contexts in the past five years.Findings(1) China’s research in the field of information behavior started relatively late, but the volume of publications has grown rapidly since 2004, currently ranking second globally in cumulative publication quantity. However, the influence of the literature published by China is limited, and there is a lack of research institutions with global influence. (2) In the last five years, China has conducted extensive research in various information contexts. Among these, most research was conducted in work contexts, followed by healthcare contexts, especially studies related to epidemics. (3) Current research on information behavior in China is characterized by expanded and refined research groups, diversified research categories, continuous expansion and enrichment of research contexts, increased interdisciplinary nature of research and continuous innovation in research methods and theoretical models.Originality/valueThis study, utilizing a scientific knowledge map, elucidates China’s position in global information behavior research, with a specific emphasis on analyzing China’s research hot topics and trends in this field over the past five years. It aims to provide valuable resources for scholars interested in understanding the status of information behavior research in China and to offer some guidance for scholars currently or intending to engage in information behavior research.

DFSTrack: Dual-stream fusion Siamese network for human pose tracking in videos

Human pose tracking is a challenging task that involves estimating the human pose and tracking it across multiple frames in a video sequence. In recent years, deep learning-based methods have made significant progress in this field, achieving state-of-the-art performance. However, due to complex background and occlusion among people missed detection and incorrect association matching are still the challenging problems. To address these issues, we adopt a top-down framework to perform human pose tracking in the paper. We propose a human detection prediction recovery module (HDP module) to recover missed detection, and propose a dual-stream fusion Siamese network for human matching (DFSTrack). Specifically, we design a residual graph convolutional block (RGCN block) for spatial position encoding of human keypoints, and use spatial self-attention and temporal cross-attention to design a dual-stream spatial–temporal fusion transformer (DST Transformer). The graph convolutional block and transformer are cascaded to simultaneously obtain information on the spatial and temporal positions of human keypoints, allowing the Siamese network to solve the erroneous human matching. Experimental results on the PoseTrack17 dataset, PoseTrack18 dataset and PoseTrack21 dataset demonstrate that our proposed method outperforms state-of-the-art methods on human pose tracking tasks. Our code and pretrained models are available at https://github.com/yhtian2023/DFSTrack.

Spectral recomposition feature for optimizing seismic velocity model prediction with a neural network

An approach incorporating spectral recomposition results for training a neural network is proposed. The initial procedure is based on an inversion scheme for reconstructing the seismic spectrum of wavelets associated with a reflection event. This allows one to estimate each wavelet's temporal position within a seismogram. Since each calculated wavelet's temporal position is identified, this information is integrated into the seismogram as an additional feature for a trained neural network. A Fully Convolutional Network is trained with models that have features commonly found in offshore hydrocarbon reservoirs. The training step is based on composing a set of an acoustic velocity model set with its seismograms and their position matrices at the time of each wavelet for each seismogram. Next, the network training with a specific number of sets is performed. This approach was able to predict acoustic velocity models more accurately, especially for predicting the shape of salt bodies and the interface of inclined layers, which makes the proposed approach an interesting mean for estimating specific features in offshore hydrocarbon reservoir structures.

Evaluation of the cavotricuspid isthmus contractile kinetics within an interventional CMR environment

Abstract Background Cavotricuspid isthmus (CTI)-dependent flutter is a common arrhythmia with recurrences following ablation up to 50%. Previous static morphological analysis have shown that CTI anatomical variability has an impact on ablation time and procedural success. However, whether the contractile kinetics of the CTI affect procedural success remains unknown to date. Purpose Assessment of the contractile kinetics of the CTI in a cardiac magnetic resonance (CMR) environment for patients (pts) referred for typical flutter ablation. Method 32 consecutive pts with typical right atrial (RA) flutter from our Heart Center were included in this study. RA wall contours during a single cardiac cycle were delineated on CMR 2D-images using the Tomtec software. CTIs were divided in 5 segments for detailed motion analysis as shown in the figure 1A. CTI length and displacement were measured both in caudo-cranial and postero-anterior directions. Quantification of the CTI kinetics was performed in Python by computing the 2D-Euclidean distance (between two extreme temporal positions of CTI points). Results The mean CTI length of the study population (mean age 66±10 years, 29 males) was 32±13 mm during RA systole and 49±15 mm during RA diastole, resulting in a mean CTI change of 17±7 mm over one cardiac cycle (maximum – minimum length). Figure 1A shows representative MRI views of the CTI during RA systole and diastole. Figure 1B shows examples of hypo- and hyperkinetics CTI from 2 representative pts. Figure 2C shows a gradual increase in segment displacement postero-anteriorly (8.1±7.3 vs 17.2±8.1 mm, p&amp;lt;0.05) and caudo-cranially (1.6±0.8 vs 4.5±2.4 mm) from the Eustachian valve toward the Tricuspid valve. Figure 2D indicates that the 2D-Euclidean displacement of the 5 segments varied between 8±7.7 mm and 18±8.5 mm depending on their location. Interestingly, the overall CTI displacement was 4.5 times greater forward than upward. Conclusion To our knowledge, this is the first study investigating the kinetics properties of the CTI in sinus rhythm of patients with typical RA flutter. The CTI showed displacements both in postero-anterior and caudo-cranial directions, with a gradual increase from the Eustachian to the Tricuspid valves. CTI displacement appeared highly variable across patients. Whether the kinetics of the CTI correlate with ablation success will be investigated shortly.

Oceanic feeling in Mati Diop’s Atlantique

ABSTRACT The ocean is a constant presence in Mati Diop’s award-winning film Atlantique/Atlantics (2019), a ghostly love story about the ongoing migration crisis set in Dakar, Senegal. A critical issue Diop explores in this film is the contradiction between the geopolitical stasis of a high-rise tower and the expansive continuity of the sea, making the physical reality captured by the lens overlap with the unrestricted flow of presences carried by waves of sea and sound through the film frame. This article analyses oceanic aesthetics and the unrestricted flows of bodies of water and work in both Diop’s recent feature film and her first experimental documentary of almost the same title, Atlantiques/Atlantics (2009). These corporeal drifts disrupt narrative and temporal flows and position the cinema as a space for poetic and political change. What the ocean embodies is not only a tragic archive of colonial violence, as Rosalind Galt has recently argued, but also a space for a feminist and decolonial refusal of such loss.