Motion Dataset Research Articles

Home-Based Real-Time Abnormal Movement Detection System Deployed on On-Device Artificial Intelligence

In response to the aging trend in society and to Human Augmentation beings for home-based activities, this paper proposes an Abnormal Movement Detection system, using the common at-home movements of standing up and hand tremors while picking up items for abnormal movement verification. This can be easily applied in ordinary homes or long-term care institutions; for those living alone with limited resources, there is no longer any need to purchase expensive monitoring equipment to achieve improved quality of life. Therefore, this research collected and built the own dataset as the first important step of the study. The proposed Abnormal Movement Detection system is implemented by designing a deep learning network. Several issues, including the network architecture, the novel method of data augmentation and the scoring method of expanding the intervals between abnormality levels, are studied. For achieving the home-based real-time detection, there are four main contributions of this paper. The first is that a training dataset was collected and established: From this, the pathognomonic movement categories are easy to observe in home activities and geometric data augmentation can be used to improve the related home activity video collection. The second is the abnormal behavior detection architecture: This architecture has several important function blocks including detecting object, detecting action, inspecting abnormal movement and reminding event, using Convolutional Neural Network combined with Long Short-Term Memory ([Formula: see text]) as the core network for abnormal motion detection. With movement abnormality evaluation based on different levels, it can judge abnormal behaviors and conduct model training, performance evaluation and architecture optimization with both public domain datasets and the movement dataset collected in this research project. The third is the proliferation of new attributes in the videos: New attributes are added to the original videos through a Generative Adversarial Network (GAN), producing new training videos; the effectiveness of two different generation methods is evaluated. Finally, the algorithms developed in this paper are deployed on resource-constrained On-device Artificial Intelligence (AI). Activity videos from a total of 20 people were collected; in all, 53 videos of StandUp and 60 videos of PickUpItems were obtained to establish the training dataset. When CNN and LSTM network were added to Batch Normalization (BN), and Global Average Pooling (GAP) replaced Fully Connected (FC) layers, the accuracy rate reached 98.4%. In terms of data augmentation, geometric transformations and GAN were used to estimate the performance. The experimental results showed that the geometric transformation using brightness adjustment had the highest accuracy rate of 98.6%. Finally, the Softmax layer using Phi-Softmax–tan(⋅) function was shown to be the best method to expand the intervals between abnormality levels.

Experimental Study: Deep Learning-Based Fall Monitoring among Older Adults with Skin-Wearable Electronics.

Older adults are more vulnerable to falling due to normal changes due to aging, and their falls are a serious medical risk with high healthcare and societal costs. However, there is a lack of automatic fall detection systems for older adults. This paper reports (1) a wireless, flexible, skin-wearable electronic device for both accurate motion sensing and user comfort, and (2) a deep learning-based classification algorithm for reliable fall detection of older adults. The cost-effective skin-wearable motion monitoring device is designed and fabricated using thin copper films. It includes a six-axis motion sensor and is directly laminated on the skin without adhesives for the collection of accurate motion data. To study accurate fall detection using the proposed device, different deep learning models, body locations for the device placement, and input datasets are investigated using motion data based on various human activities. Our results indicate the optimal location to place the device is the chest, achieving accuracy of more than 98% for falls with motion data from older adults. Moreover, our results suggest a large motion dataset directly collected from older adults is essential to improve the accuracy of fall detection for the older adult population.

Tax Planning Knowledge Diffusion via the Labor Market

We examine the extent to which the labor market facilitates the diffusion of tax-planning knowledge across firms. Using a novel data set of tax department employee movements between S&P 1500 firms, we find that firms experience an increase in their tax planning after hiring a tax employee from a tax-aggressive firm. This finding is robust to various research designs and specifications. Consistent with tax-planning knowledge driving the result, we find that the tax-planning benefits are more substantial when the employee is involved in a director-level role and has more experience. Further tests suggest that tax-planning knowledge is highly specific in nature: the increase in tax avoidance is larger when the hiring and former firms are similar (i.e., operating in the same sector or having similar foreign operations), and firms are more likely to hire tax department employees from firms with similar characteristics. Finally, we do not find that the prior firm’s tax planning changes after the employee leaves the firm, suggesting that the tax-planning knowledge simply spreads to the hiring firm and does not leave the prior firm. Our study documents the first order role of the labor market in the diffusion of tax-planning knowledge across firms, and our findings suggest that tax department human capital is a central determinant of tax-planning outcomes. This paper was accepted by Suraj Srinivasan, accounting. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2023.4741 .

A three-stage multi-objective heterogeneous integrated model with decomposition-reconstruction mechanism and adaptive segmentation error correction method for ship motion multi-step prediction

Reliable and accurate ship motion prediction is essential for ship navigation at sea and marine operations. Although previous studies have yielded rich results in the field of ship motion prediction, most of them have ignored the importance of the dynamic characteristics of ship motion for constructing forecasting models. Besides, the limitations of the single model and the autocorrelation characteristics of the residual series are also unfavorable factors that hinder the forecasting performance. To fill these gaps, a multi-objective heterogeneous integration model based on decomposition-reconstruction mechanism and adaptive segmentation error correction method is proposed in this paper for ship motion multi-step prediction. Specifically, the proposed model is divided into three stages, which are decomposition-reconstruction mechanism, multi-objective heterogeneous integration model and adaptive segmentation error correction method. The effectiveness of the proposed model is verified using four sets of real ship motion data collected from two sites in the South China Sea. The evaluation results show that the proposed model can effectively improve the prediction performance and outperforms other traditional models and state-of-the-art models in the field of ship motion prediction. Prospectively, the model proposed in this study can be used as an effective aid to ship warning systems and has the potential for practical application in ship marine operations.

Reconstruction of physical dance teaching content and movement recognition based on a machine learning model

With the technological development of movement recognition based on machine learning model algorithms, the content and movements for physical dance teaching are also seeking changes and innovations. In this paper, a set of three-dimensional convolutional neural network recognition algorithms based on a machine learning model is constructed through the collection to recognition of sports dance movement data. By collecting the skeleton information of typical movements of physical dance, a typical movement dataset of physical dance is constructed, which is recognized by the improved 3D convolutional neural network recognition algorithm under the machine learning model, and the method is validated on the public dataset. The experimental results show that the 3D CNNs in this paper can produce relatively satisfactory results for sports dance action recognition with high accuracy of action recognition, which verifies the feasibility of the 3D convolutional neural network action recognition algorithm under the machine learning model for the acquisition to recognition of sports dance actions. It illustrates that the future can be better to open a new direction of physical dance education content through machine learning models in this form.

A dataset of asymptomatic human gait and movements obtained from markers, IMUs, insoles and force plates

Human motion capture and analysis could be made easier through the use of wearable devices such as inertial sensors and/or pressure insoles. However, many steps are still needed to reach the performance of optoelectronic systems to compute kinematic parameters. The proposed dataset has been established on 10 asymptomatic adults. Participants were asked to walk at different speeds on a 10-meters walkway in a laboratory and to perform different movements such as squats or knee flexion/extension tasks. Three-dimensional trajectories of 69 reflective markers placed according to a conventional full body markerset, acceleration and angular velocity signals of 8 inertial sensors, pressure signals of 2 insoles, 3D ground reaction forces and moments obtained from 3 force plates were simultaneously recorded. Eight calculated virtual markers related to joint centers were also added to the dataset. This dataset contains a total of 337 trials including static and dynamic tasks for each participant. Its purpose is to enable comparisons between various motion capture systems and stimulate the development of new methods for gait analysis.

Detection of Movement Related Cortical Potentials from Single Trial EEG Signals

Movement-Related Cortical Potentials (MRCP) are signals that begin to appear approximately two seconds before the onset of voluntary movements and can be recorded with EEG. MRCP is an important sign that the movement will begin. Determining the movement intention before the action is extremely important information especially for real-time BCI systems. By using MRCP, Brain-Computer Interface (BCI) users' movement intention can be determined prior to the move and this sign can be used as a control signal. In this study, it was aimed to determine the movement and resting states with high accuracy with MRCP signals. Furthermore, the effects of filter cutoff frequencies, number of electrodes, and MRCP time interval window on the success of distinguishing movement/resting states in the preprocessing stage were investigated. For this purpose, Katz fractal dimension and nonlinear support vector machine methods were used in the feature extraction and classification stages, respectively. The proposed method was tested on the attempted hand and arm movements dataset containing EEG signals of 10 participants with spinal cord injury. Katz fractal dimension and support vector machines methods can determine movement and resting states with an average of 96.47% accuracy using MRCP signals. If the number of electrodes to be used in signal analysis was 3, 9 and 61, the obtained accuracy rates were determined as 83.71%, 90.67%, and 96.47%, respectively. The experimental results also showed that the filter cutoff frequencies used in the preprocessing had a significant effect on the accuracy.

Swin-TCNet: Swin-based temporal-channel cascade network for motor imagery iEEG signal recognition

Intracranial electroencephalography (iEEG) can record neuronal activities in different brain areas by placing strip or mesh electrodes on the surface of the brain. Benefiting from the high temporal and spatial resolution of iEEG signals, iEEG-based motor imagery has gained great attention, and made significant progress under the frameworks of convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, both CNN- and RNN-based methods have limitations in perceiving the long-range global dependencies of local features in time and spatial dimensions. To exploit the temporal and spatial relationship of iEEG signals, we propose a Swin-Based Temporal Cascade Channel Network (Swin-TCNet) for motor imagery classification tasks, which is composed of three modules: Temporal-Swin (TS), Channel-Swin (CS) and Classifier. The TS and CS modules respectively extract the global dependencies of iEEG signals in the time and channel dimensions, and the Classifier assigns the class labels to the inputs. In addition, we introduce a Temporal-Channel Swin (TC-Swin) block instead of the standard Swin block to effectively reduce the computational complexity. We validate our model on the public imagined handwriting movement dataset collected by Stanford University. The experimental results demonstrate that our model can raise the classification accuracy by 2%–9% compared with 6 state-of-the-art models.

Strong motion estimation in Costa Rica at non-record sites using spectral inversion method

We performed the spectral inversion to separate the source, path, and site effects using strong ground motion records from essential facilities such as powerhouses and hydroelectric dams, distributed all over Costa Rica. The spectral inversion technique was adapted to a sparsely observed strong motion data set. We used the results to estimate strong ground motions at sites where there were no observation records in and around the intensity anomaly that appeared during the 2012 Sámara earthquake of Mw 7.6. To ensure the quality, we incorporated a broadband station to estimate levels of acceleration source spectra for six events 5.3 ≤ Mw ≤ 6.5, using them as the reference events for the spectral inversion. In this paper, we introduced two consistency checks: confirmation of the correctness of the input data, and verification of the output using synthetic spectra. We used them as an iterative implementation of sequences of elimination of suspicious data and inversions. We regulated the source spectra of the reference events for the final inversion, considering the estimate of site effect for one-dimensional SH wave propagation at the site of the lowest site amplification. We obtained the frequency-dependent quality factor Q = 179f0.5598 for the northern and central parts of Costa Rica. From the output of the spectral inversion, we reproduced the acceleration spectra of earthquakes in the sites where the event was not recorded. We applied this formulation using the earthquake mentioned above. The resulting synthetic spectra are consistent with the anomalous intensity distribution reported at that time.

Investigating the Accuracy of Forecasting Travel Duration for Uber Trips and Analyzing the Influence of Weather on Traffic Flow

Accurately predicting the travel time between two destinations is an essential aspect of traffic monitoring and facilitating ridesharing services. However, this is a highly complex and challenging task, which involves a multitude of variables that cannot be resolved straightforwardly. Previous studies on travel time prediction have focused on evaluating the duration of individual road segments or specific sub-paths before integrating the necessary time for each sub-path. While this method may provide some insight, it may result in an incorrect or imprecise time estimate. To address this issue, this research aims to utilize machine learning techniques to predict the duration of trips in ride-sharing networks, by utilizing the Uber movement dataset. The proposed system employs Python programming to calculate the distance between the pickup and drop-off locations. Furthermore, the study explores the various factors that affect travel time in a descriptive analysis. This includes examining the impact of traffic congestion, weather conditions, and road construction on travel time. The suggested approach incorporates a robust regression model known as Huber regression to enhance the accuracy of trip duration prediction and increase the precision of the algorithm. The Huber regression model is robust to outliers, making it suitable for the Uber movement dataset, which may contain unexpected and extreme values. The dataset is processed using k-fold cross-validation, which splits the dataset into k subsets, with each subset used for validation once while the remaining subsets used for training the model. However, this approach presents several challenges that need to be addressed, including the difficulties with tracking variables, the need for extensive data transformation due to the diverse data types contained in the dataset, and the challenge of handling unlabeled places during the segmentation of geographical data. Additionally, outliers in the dataset can lead to substantial data differences and affect the model's accuracy. Data normalization is slow due to the time-consuming nature of reading duplicated information. To mitigate these issues, additional study is required to improve the model's layout and address the challenges of working with the Uber movement dataset.

Deep Learning-Based Motion Style Transfer Tools, Techniques and Future Challenges

In the fourth industrial revolution, the scale of execution for interactive applications increased substantially. These interactive and animated applications are human-centric, and the representation of human motion is unavoidable, making the representation of human motions ubiquitous. Animators strive to computationally process human motion in a way that the motions appear realistic in animated applications. Motion style transfer is an attractive technique that is widely used to create realistic motions in near real-time. motion style transfer approach employs existing captured motion data to generate realistic samples automatically and updates the motion data accordingly. This approach eliminates the need for handcrafted motions from scratch for every frame. The popularity of deep learning (DL) algorithms reshapes motion style transfer approaches, as such algorithms can predict subsequent motion styles. The majority of motion style transfer approaches use different variants of deep neural networks (DNNs) to accomplish motion style transfer approaches. This paper provides a comprehensive comparative analysis of existing state-of-the-art DL-based motion style transfer approaches. The enabling technologies that facilitate motion style transfer approaches are briefly presented in this paper. When employing DL-based methods for motion style transfer, the selection of the training dataset plays a key role in the performance. By anticipating this vital aspect, this paper provides a detailed summary of existing well-known motion datasets. As an outcome of the extensive overview of the domain, this paper highlights the contemporary challenges faced by motion style transfer approaches.

A fast nonlinear dynamic analysis automated approach to produce fragility curves for 3D RC frames

This study describes an automated procedure to assess seismic fragility of 3D reinforced concrete frame structures. A number of 9600 3D nonlinear dynamic analyses are performed using the recently developed 3D Force Analogy Method, implemented in an automated Matlad-based non-commercial code. The numerical efficiency of 3D Force Analogy is first illustrated for a broad range of structures and groud motion recordings and scaling factors. The ground motion dataset consisting of 20 horizontal recordings used in the study is characteristic for soft deep soil sites (e.g Bucharest) which are affected by large magnitude earthquakes originating in the Vrancea intermediate-depth seismic source. A total of 24 RC frame structures, both regular and irregular are employed in the study. The Park-Ang damage index is employed in order to evaluate the damage states of the structures. Fragility functions corresponding to life-safety and collapse damage states are obtained for both principal directions of the structures. The results show similar median fragility values on both principal directions, and a larger variability on the transversal direction. The analyses have also shown that the standard deviations evaluated for the mean annual failure rates corresponding to the two considered damage states and on both principal directions is almost equal to the mean value. The mean value of the collapse probability evaluated for the peak ground acceleration with a mean return period of 2475 years and based on the results of all 24 models is 6.1 % for the longitudinal direction and 6.5 % for the transversal direction.

Benefits of Shared-Fleet Horizontal Logistics Collaborations: A Case Study of Patient Service Vehicles Collecting Pathology Samples in a Public Sector Healthcare Setting

Road-based logistics suffer from inefficiencies due to less-than-full load vehicle movements. Consolidating loads through shared-fleet collaborations (also known as freight pooling) can reduce such inefficiencies, and thereby reduce costs, vehicle-kilometres (vkm), and related emissions and congestion. Utilising a significant historical dataset of vehicle movements, the potential cost savings and environmental benefits of a shared-fleet operation involving collaboration between two public sector organisations, integrating both static (fixed-schedule) and dynamic (client-specific) demand within a healthcare setting, were quantified. A Sample Collection Service (SCS; transporting pathology samples from doctors’ surgeries to centralised laboratories for analysis) shared spare capacity in vehicles operated by a Patient Transport Service (PTS; transporting eligible non-emergency patients to/from routine hospital appointments) as an alternative to engaging an external courier company. Results suggested that a shared-fleet collaboration servicing 78 surgeries, alongside normal patient loads in an average of 24 PTS vehicles/day, produced reductions of 16%, 13% and 12% in costs, vkm and carbon dioxide emissions, respectively. Decision-makers within public sector organisations that operate own-account vehicle fleets could pursue policies that actively seek out opportunities to deploy shared-fleet solutions to improve vehicle utilisation and therefore reduce public sector spending and the detrimental effects of road logistics.

Motion prediction for beating heart surgery with GRU

This work aims to * predict * the * 3D * coordinates of the * point * of * interest * (POI) on * the surface of beating heart in dynamic minimally invasive surgery, which can improve the manoeuvrability of cardiac surgical robots and expand their functions. For accurate and robust POI motion prediction, a deep learning technique, the gated recurrent unit (GRU), is employed to learn the spatio-temporal (ST) correlation of the POI and its auxiliary points (APs) from their past trajectories. For reference, two neural network models that exploit only spatial and temporal correlations, respectively, are also investigated. The prediction accuracy and robustness of the above three models are verified based on the motion datasets of phantom and in vivo hearts collected by da Vinci robots. Source codes and motion data are shared on GitHub (https://github.com/oww-file/ST-GRU). The experimental results show that the proposed ST-GRU model is significantly better than the other two reference models, and it outperforms the state-of-the-art deep learning method on the same datasets.

Describing movement learning using metric learning.

Analysing movement learning can rely on human evaluation, e.g. annotating video recordings, or on computing means in applying metrics on behavioural data. However, it remains challenging to relate human perception of movement similarity to computational measures that aim at modelling such similarity. In this paper, we propose a metric learning method bridging the gap between human ratings of movement similarity in a motor learning task and computational metric evaluation on the same task. It applies metric learning on a Dynamic Time Warping algorithm to derive an optimal set of movement features that best explain human ratings. We evaluated this method on an existing movement dataset, which comprises videos of participants practising a complex gesture sequence toward a target template, as well as the collected data that describes the movements. We show that it is possible to establish a linear relationship between human ratings and our learned computational metric. This learned metric can be used to describe the most salient temporal moments implicitly used by annotators, as well as movement parameters that correlate with motor improvements in the dataset. We conclude with possibilities to generalise this method for designing computational tools dedicated to movement annotation and evaluation of skill learning.

A Music-Driven Deep Generative Adversarial Model for Guzheng Playing Animation.

To date relatively few efforts have been made on the automatic generation of musical instrument playing animations. This problem is challenging due to the intrinsically complex, temporal relationship between music and human motion as well as the lacking of high quality music-playing motion datasets. In this article, we propose a fully automatic, deep learning based framework to synthesize realistic upper body animations based on novel guzheng music input. Specifically, based on a recorded audiovisual motion capture dataset, we delicately design a generative adversarial network (GAN) based approach to capture the temporal relationship between the music and the human motion data. In this process, data augmentation is employed to improve the generalization of our approach to handle a variety of guzheng music inputs. Through extensive objective and subjective experiments, we show that our method can generate visually plausible guzheng-playing animations that are well synchronized with the input guzheng music, and it can significantly outperform the state-of-the-art methods. In addition, through an ablation study, we validate the contributions of the carefully-designed modules in our framework.

Can a Computational Model Predict the Effect of Lesion Location on Cam-type Hip Impingement?

The Warwick consensus defined femoroacetabular impingement syndrome as a motion-related clinical disorder of the hip with a triad of symptoms, clinical signs, and imaging findings representing symptomatic premature contact between the proximal femur and acetabulum. Several factors appear to cause labral and cartilage damage, including joint shape and orientation and patient activities. There is a lack of tools to predict impingement patterns in a patient across activities. Current computational modeling tools either measure pure ROM of the joint or include complexity that reduces reliability and increases time to achieve a solution. The purpose of this study was to examine the efficacy of a low computational cost approach to combining cam-type hip shape and multiple hip motions for predicting impingement. Specifically, we sought to determine (1) the potential to distinguish impingement in individual hip shapes by analyzing the difference between a cam lesion at the anterior femoral neck and one located at the superior femoral neck; (2) sensitivity to three aspects of hip alignment, namely femoral neck-shaft angle, femoral version angle, and pelvic tilt; and (3) the difference in impingement measures between the individual activities in our hip motion dataset. A model of the shape and alignment of a cam-type impinging hip was created and used to describe two locations of a cam lesion on the femoral head-neck junction (superior and anterior) based on joint shape information available in prior studies. Sensitivity to hip alignment was assessed by varying three aspects from a baseline (typical alignment described in prior studies), namely, femoral neck-shaft angle, femoral version, and pelvic tilt. Hip movements were selected from an existing database of 18 volunteers performing 13 activities (10 male, eight female; mean age 44 ± 19 years). A subset was selected to maximize variation in the range of joint angles and maintain a consistent number of people performing each activity, which resulted in nine people per activity, including at least three of each sex. Activities included pivoting during walking, squatting, and golf swing. All selected hip motion cases were applied to each hip shape model. For the first part of the study, the number of motion cases in which impingement was predicted was recorded. Quantitative analyses of the depth of penetration of the cam lesion into the acetabular socket and qualitative observations of impingement location were made for each lesion location (anterior and superior). In the second part of the study, in which we aimed to test the sensitivity of the findings to hip joint orientation, full analysis of both cam lesion locations was repeated for three modified joint orientations. Finally, the results from the first part of the analysis were divided by activity to understand how the composition of the activity dataset affected the results. The two locations of cam lesion generated impingement in a different percentage of motion cases (anterior cam: 56% of motion cases; superior cam: 13% of motion cases) and different areas of impingement in the acetabulum, but there were qualitatively similar penetration depths (anterior cam: 6.8° ± 5.4°; superior cam: 7.9° ± 5.8°). The most substantial effects of changing the joint orientation were a lower femoral version angle for the anterior cam, which increased the percentage of motion cases generating impingement to 67%, and lower neck-shaft angle for the superior cam, which increased the percentage of motion cases generating impingement to 37%. Flexion-dominated activities (for example, squatting) only generated impingement with the anterior cam. The superior cam generated impingement during activities with high internal-external rotation of the joint (for example, the golf swing). This work demonstrated the capability of a simple, rapid computational tool to assess impingement of a specific cam-type hip shape (under 5 minutes for more than 100 motion cases). To our knowledge, this study is the first to do so for a large set of motion cases representing a range of activities affecting the hip, and could be used in planning surgical bone removal. The results of this study imply that patients with femoroacetabular impingement syndrome with cam lesions on the superior femoral head-neck junction may experience impinging during motions that are not strongly represented by current physical diagnostic tests. The use of this tool for surgical planning will require streamlined patient-specific hip shape extraction from imaging, model sensitivity testing, evaluation of the hip activity database, and validation of impingement predictions at an individual patient level.

Quantified analysis of facial movement: A reference for clinical applications.

Most techniques for evaluating unilateral impairments in facial movement yield subjective measurements. The objective of the present study was to define a reference dataset and develop a visualization tool for clinical assessments. In this prospective study, a motion capture system was used to quantify facial movements in 30 healthy adults and 2 patients. We analyzed the displacements of 105 reflective markers placed on the participant's face during five movements (M1-M5). For each marker, the primary endpoint was the maximum amplitude of displacement from the static position (M0) in an analysis of variance. The measurement precision was 0.1 mm. Significant displacements of markers were identified for M1-M5, and displacement patterns were defined. The patients and age-matched healthy participants were compared with regard to the amplitude of displacement. We created a new type of radar plot to visually represent the diagnosis and facilitate effective communication between medical professionals. In proof-of-concept experiments, we collected quantitative data on patients with facial palsy and created a patient-specific radar plot. Our new protocol for clinical facial motion capture ("quantified analysis of facial movement," QAFM) was accurate and should thus facilitate the long-term clinical follow-up of patients with facial palsy. To take account of the limitations affecting the comparison with the healthy side, we created a dataset of healthy facial movements; our method might therefore be applicable to other conditions in which movements on one or both sides of the face are impaired. The patient-specific radar plot enables clinicians to read and understand the results rapidly.

Rapid absolute plate motion changes inferred from high-resolution relative spreading reconstructions: A case study focusing on the South America plate and its Atlantic/Pacific neighbors

The reconstruction of past plate motions relative to a deemed–to–be–fixed hotspot reference frame relies on the sparse occurrence of intraplate volcanism. Consequently, this absolute reference frame often features a temporal resolution that exceeds the rapid kinematic changes observed in plate–to–plate spreading reconstructions, changes recently shown to occur within less than 5 Ma. In this work we put forward an alternative method based on the study of high–resolution relative plate motion data sets. By studying time periods featuring a relatively high probability of plate motion change across multiple spreading ridges, we are able to identify and quantify (likely) changes in absolute plate motion. Specifically, we implement such approach and provide well–defined estimates for the absolute plate motion changes of South America and neighboring plates Nubia, Antarctica, Somalia, North America and Pacific. We find that kinematic changes for all these plates occur between 9 and 5 Ma. For South America, we identify a change also between 14 and 10 Ma. Lastly, we estimate the torque–variations required upon these plates to generate the inferred kinematic changes, which we find to be between ∼5⋅1023 and ∼20⋅1024 N ⋅ m.

Mesolimbic dopamine adapts the rate of learning from action

Recent success in training artificial agents and robots derives from a combination of direct learning of behavioural policies and indirect learning through value functions1–3. Policy learning and value learning use distinct algorithms that optimize behavioural performance and reward prediction, respectively. In animals, behavioural learning and the role of mesolimbic dopamine signalling have been extensively evaluated with respect to reward prediction4; however, so far there has been little consideration of how direct policy learning might inform our understanding5. Here we used a comprehensive dataset of orofacial and body movements to understand how behavioural policies evolved as naive, head-restrained mice learned a trace conditioning paradigm. Individual differences in initial dopaminergic reward responses correlated with the emergence of learned behavioural policy, but not the emergence of putative value encoding for a predictive cue. Likewise, physiologically calibrated manipulations of mesolimbic dopamine produced several effects inconsistent with value learning but predicted by a neural-network-based model that used dopamine signals to set an adaptive rate, not an error signal, for behavioural policy learning. This work provides strong evidence that phasic dopamine activity can regulate direct learning of behavioural policies, expanding the explanatory power of reinforcement learning models for animal learning6.