Parkinson's disease (PD) is a common neurodegenerative disorder with a prevalence rate estimated to 2.0% for people aged over 65 years. Cardinal motor symptoms of PD such as rigidity and bradykinesia affect the muscles involved in the handwriting process resulting in handwriting abnormalities called PD dysgraphia. Nowadays, online handwritten signal (signal with temporal information) acquired by the digitizing tablets is the most advanced approach of graphomotor difficulties analysis. Although the basic kinematic features were proved to effectively quantify the symptoms of PD dysgraphia, a recent research identified that the theory of fractional calculus can be used to improve the graphomotor difficulties analysis. Therefore, in this study, we follow up on our previous research, and we aim to explore the utilization of various approaches of fractional order derivative (FD) in the analysis of PD dysgraphia. For this purpose, we used the repetitive loops task from the Parkinson's disease handwriting database (PaHaW). Handwritten signals were parametrized by the kinematic features employing three FD approximations: Gr\"unwald-Letnikov's, Riemann-Liouville's, and Caputo's. Results of the correlation analysis revealed a significant relationship between the clinical state and the handwriting features based on the velocity. The extracted features by Caputo's FD approximation outperformed the rest of the analyzed FD approaches. This was also confirmed by the results of the classification analysis, where the best model trained by Caputo's handwriting features resulted in a balanced accuracy of 79.73% with a sensitivity of 83.78% and a specificity of 75.68%.
The ability to generate synthetic sequences is crucial for a wide range of applications, and recent advances in deep learning architectures and generative frameworks have greatly facilitated this process. Particularly, unconditional one-shot generative models constitute an attractive line of research that focuses on capturing the internal information of a single image, video, etc. to generate samples with similar contents. Since many of those one-shot models are shifting toward efficient non-deep and non-adversarial approaches, we examine the versatility of a one-shot generative model for augmenting whole datasets. In this work, we focus on how similarity at the subsequence level affects similarity at the sequence level, and derive bounds on the optimal transport of real and generated sequences based on that of corresponding subsequences. We use a one-shot generative model to sample from the vicinity of individual sequences and generate subsequence-similar ones and demonstrate the improvement of this approach by applying it to the problem of Unmanned Aerial Vehicle (UAV) identification using limited radio-frequency (RF) signals. In the context of UAV identification, RF fingerprinting is an effective method for distinguishing legitimate devices from malicious ones, but heterogenous environments and channel impairments can impose data scarcity and affect the performance of classification models. By using subsequence similarity to augment sequences of RF data with a low ratio (5\%-20\%) of training dataset, we achieve significant improvements in performance metrics such as accuracy, precision, recall, and F1 score.
Uncertainty quantification is crucial to inverse problems, as it could provide decision-makers with valuable information about the inversion results. For example, seismic inversion is a notoriously ill-posed inverse problem due to the band-limited and noisy nature of seismic data. It is therefore of paramount importance to quantify the uncertainties associated to the inversion process to ease the subsequent interpretation and decision making processes. Within this framework of reference, sampling from a target posterior provides a fundamental approach to quantifying the uncertainty in seismic inversion. However, selecting appropriate prior information in a probabilistic inversion is crucial, yet non-trivial, as it influences the ability of a sampling-based inference in providing geological realism in the posterior samples. To overcome such limitations, we present a regularized variational inference framework that performs posterior inference by implicitly regularizing the Kullback-Leibler divergence loss with a CNN-based denoiser by means of the Plug-and-Play methods. We call this new algorithm Plug-and-Play Stein Variational Gradient Descent (PnP-SVGD) and demonstrate its ability in producing high-resolution, trustworthy samples representative of the subsurface structures, which we argue could be used for post-inference tasks such as reservoir modelling and history matching. To validate the proposed method, numerical tests are performed on both synthetic and field post-stack seismic data.
Compared to regular cameras, Dynamic Vision Sensors or Event Cameras can output compact visual data based on a change in the intensity in each pixel location asynchronously. In this paper, we study the application of current image-based SLAM techniques to these novel sensors. To this end, the information in adaptively selected event windows is processed to form motion-compensated images. These images are then used to reconstruct the scene and estimate the 6-DOF pose of the camera. We also propose an inertial version of the event-only pipeline to assess its capabilities. We compare the results of different configurations of the proposed algorithm against the ground truth for sequences of two publicly available event datasets. We also compare the results of the proposed event-inertial pipeline with the state-of-the-art and show it can produce comparable or more accurate results provided the map estimate is reliable.
Conversational AI has become an increasingly prominent and practical application of machine learning. However, existing conversational AI techniques still suffer from various limitations. One such limitation is a lack of well-developed methods for incorporating auxiliary information that could help a model understand conversational context better. In this paper, we explore how persona-based information could help improve the quality of response generation in conversations. First, we provide a literature review focusing on the current state-of-the-art methods that utilize persona information. We evaluate two strong baseline methods, the Ranking Profile Memory Network and the Poly-Encoder, on the NeurIPS ConvAI2 benchmark dataset. Our analysis elucidates the importance of incorporating persona information into conversational systems. Additionally, our study highlights several limitations with current state-of-the-art methods and outlines challenges and future research directions for advancing personalized conversational AI technology.
Understanding the probabilistic traffic environment is a vital challenge for the motion planning of autonomous vehicles. To make feasible control decisions, forecasting future trajectories of adjacent cars is essential for intelligent vehicles to assess potential conflicts and react to reduce the risk. This paper first introduces a Bayesian Long Short-term Memory (BLSTM) model to learn human drivers' behaviors and habits from their historical trajectory data. The model predicts the probability distribution of surrounding vehicles' positions, which are used to estimate dynamic conflict risks. Next, a hybrid automaton is built to model the basic motions of a car, and the conflict risks are assessed for real-time state-space transitions based on environmental information. Finally, a BLSTM-based Model Predictive Control (MPC) is built to navigate vehicles through safe paths with the least predicted conflict risk. By merging BLSTM with MPC, the designed neural-based MPC overcomes the defect that traditional MPC is hard to model uncertain conflict risks. The simulation results show that our proposed BLSTM-based MPC performs better than human drivers because it can foresee potential conflicts and take action to avoid them.
Local news articles are a subset of news that impact users in a geographical area, such as a city, county, or state. Detecting local news (Step 1) and subsequently deciding its geographical location as well as radius of impact (Step 2) are two important steps towards accurate local news recommendation. Naive rule-based methods, such as detecting city names from the news title, tend to give erroneous results due to lack of understanding of the news content. Empowered by the latest development in natural language processing, we develop an integrated pipeline that enables automatic local news detection and content-based local news recommendations. In this paper, we focus on Step 1 of the pipeline, which highlights: (1) a weakly supervised framework incorporated with domain knowledge and auto data processing, and (2) scalability to multi-lingual settings. Compared with Stanford CoreNLP NER model, our pipeline has higher precision and recall evaluated on a real-world and human-labeled dataset. This pipeline has potential to more precise local news to users, helps local businesses get more exposure, and gives people more information about their neighborhood safety.
Infants' neurological development is heavily influenced by their motor skills. Evaluating a baby's movements is key to understanding possible risks of developmental disorders in their growth. Previous research in psychology has shown that measuring specific movements or gestures such as face touches in babies is essential to analyse how babies understand themselves and their context. This research proposes the first automatic approach that detects face touches from video recordings by tracking infants' movements and gestures. The study uses a multimodal feature fusion approach mixing spatial and temporal features and exploits skeleton tracking information to generate more than 170 aggregated features of hand, face and body. This research proposes data-driven machine learning models for the detection and classification of face touch in infants. We used cross dataset testing to evaluate our proposed models. The models achieved 87.0% accuracy in detecting face touches and 71.4% macro-average accuracy in detecting specific face touch locations with significant improvements over Zero Rule and uniform random chance baselines. Moreover, we show that when we run our model to extract face touch frequencies of a larger dataset, we can predict the development of fine motor skills during the first 5 months after birth.
In this work, we propose a novel complementary learning approach to enhance test-time adaptation (TTA), which has been proven to exhibit good performance on testing data with distribution shifts such as corruptions. In test-time adaptation tasks, information from the source domain is typically unavailable and the model has to be optimized without supervision for test-time samples. Hence, usual methods assign labels for unannotated data with the prediction by a well-trained source model in an unsupervised learning framework. Previous studies have employed unsupervised objectives, such as the entropy of model predictions, as optimization targets to effectively learn features for test-time samples. However, the performance of the model is easily compromised by the quality of pseudo-labels, since inaccuracies in pseudo-labels introduce noise to the model. Therefore, we propose to leverage the "less probable categories" to decrease the risk of incorrect pseudo-labeling. The complementary label is introduced to designate these categories. We highlight that the risk function of complementary labels agrees with their Vanilla loss formula under the conventional true label distribution. Experiments show that the proposed learning algorithm achieves state-of-the-art performance on different datasets and experiment settings.
pyssam is a Python library for creating statistical shape and appearance models (SSAMs) for biological (and other) shapes such as bones, lungs or other organs. A point cloud best describing the anatomical 'landmarks' of the organ are required from each sample in a small population as an input. Additional information such as landmark gray-value can be included to incorporate joint correlations of shape and 'appearance' into the model. Our library performs alignment and scaling of the input data and creates a SSAM based on covariance across the population. The output SSAM can be used to parameterise and quantify shape change across a population. pyssam is a small and low dependency codebase with examples included as Jupyter notebooks for several common SSAM computations. The given examples can easily be extended to alternative datasets, and also alternative tasks such as medical image segmentation by incorporating a SSAM as a constraint for segmented organs.