In this paper we demonstrate that it is possible to generate more meaningful electroencephalography (EEG) features from raw EEG features using generative adversarial networks (GAN) to improve the performance of EEG based continuous speech recognition systems. We improve the results demonstrated by authors in [1] using their data sets for for some of the test time experiments and for other cases our results were comparable with theirs. Our proposed approach can be implemented without using any additional sensor information, whereas in [1] authors used additional features like acoustic or articulatory information to improve the performance of EEG based continuous speech recognition systems.
We introduce deep Markov spatio-temporal factorization (DMSTF), a deep generative model for spatio-temporal data. Like other factor analysis methods, DMSTF approximates high-dimensional data by a product between time-dependent weights and spatially dependent factors. These weights and factors are in turn represented in terms of lower-dimensional latent variables that we infer using stochastic variational inference. The innovation in DMSTF is that we parameterize weights in terms of a deep Markovian prior, which is able to characterize nonlinear temporal dynamics. We parameterize the corresponding variational distribution using a bidirectional recurrent network. This results in a flexible family of hierarchical deep generative factor analysis models that can be extended to perform time series clustering, or perform factor analysis in the presence of a control signal. Our experiments, which consider simulated data, fMRI data, and traffic data, demonstrate that DMSTF outperforms related methods in terms of reconstruction accuracy and can perform forecasting in a variety domains with nonlinear temporal transitions.
Given the convenience of collecting information through online services, recommender systems now consume large scale data and play a more important role in improving user experience. With the recent emergence of Graph Neural Networks (GNNs), GNN-based recommender models have shown the advantage of modeling the recommender system as a user-item bipartite graph to learn representations of users and items. However, such models are expensive to train and difficult to perform frequent updates to provide the most up-to-date recommendations. In this work, we propose to update GNN-based recommender models incrementally so that the computation time can be greatly reduced and models can be updated more frequently. We develop a Graph Structure Aware Incremental Learning framework, GraphSAIL, to address the commonly experienced catastrophic forgetting problem that occurs when training a model in an incremental fashion. Our approach preserves a user's long-term preference (or an item's long-term property) during incremental model updating. GraphSAIL implements a graph structure preservation strategy which explicitly preserves each node's local structure, global structure, and self-information, respectively. We argue that our incremental training framework is the first attempt tailored for GNN based recommender systems and demonstrate its improvement compared to other incremental learning techniques on two public datasets. We further verify the effectiveness of our framework on a large-scale industrial dataset.
In this paper, we propose Two-Stream AMTnet, which leverages recent advances in video-based action representation[1] and incremental action tube generation[2]. Majority of the present action detectors follow a frame-based representation, a late-fusion followed by an offline action tube building steps. These are sub-optimal as: frame-based features barely encode the temporal relations; late-fusion restricts the network to learn robust spatiotemporal features; and finally, an offline action tube generation is not suitable for many real-world problems such as autonomous driving, human-robot interaction to name a few. The key contributions of this work are: (1) combining AMTnet's 3D proposal architecture with an online action tube generation technique which allows the model to learn stronger temporal features needed for accurate action detection and facilitates running inference online; (2) an efficient fusion technique allowing the deep network to learn strong spatiotemporal action representations. This is achieved by augmenting the previous Action Micro-Tube (AMTnet) action detection framework in three distinct ways: by adding a parallel motion stIn this paper, we propose a new deep neural network architecture for online action detection, termed ream to the original appearance one in AMTnet; (2) in opposition to state-of-the-art action detectors which train appearance and motion streams separately, and use a test time late fusion scheme to fuse RGB and flow cues, by jointly training both streams in an end-to-end fashion and merging RGB and optical flow features at training time; (3) by introducing an online action tube generation algorithm which works at video-level, and in real-time (when exploiting only appearance features). Two-Stream AMTnet exhibits superior action detection performance over state-of-the-art approaches on the standard action detection benchmarks.
Hypertension is a medical condition that is well-established as a risk factor for many major diseases. For example, it can cause alterations in the cardiac structure and function over time that can lead to heart related morbidity and mortality. However, at the subclinical stage, these changes are subtle and cannot be easily captured using conventional cardiovascular indices calculated from clinical cardiac imaging. In this paper, we describe a radiomics approach for identifying intermediate imaging phenotypes associated with hypertension. The method combines feature selection and machine learning techniques to identify the most subtle as well as complex structural and tissue changes in hypertensive subgroups as compared to healthy individuals. Validation based on a sample of asymptomatic hearts that include both hypertensive and non-hypertensive cases demonstrate that the proposed radiomics model is capable of detecting intensity and textural changes well beyond the capabilities of conventional imaging phenotypes, indicating its potential for improved understanding of the longitudinal effects of hypertension on cardiovascular health and disease.
Epidemics, such as the present Covid-19 pandemic, usually spread at a rapid rate. Standard models, e.g., the SIR model, have stressed on the importance of finding the susceptible cases to flatten the growth rate of the spread of infection as early as possible. In the present scientific world, location traces in the form of GPS coordinates are logged by mobile device manufacturing and their operating systems developing companies, such as Apple, Samsung, Google etc. However, due to the sensitive nature of this data, it is usually not shared with other organisations, mainly to protect individual privacy. However, in disaster situations, such as epidemics, data in the form of location traces of a community of people can potentially be helpful to proactively locate susceptible people from the community and enforce quarantine on them as early as possible. Since procuring such data for the purpose of restricted use is difficult (time-consuming) due to the sensitive nature of the data, a strong case needs to be made that how could such data be useful in disaster situations. The aim of this article is to to demonstrate a proof-of-the-concept that with the availability of massive amounts of real check-in data, it is feasible to develop a scalable system that is both effective (in terms of identifying the susceptible people) and efficient (in terms of the time taken to do so). We believe that this proof-of-the-concept will encourage sharing (with restricted use) of such sensitive data in order to help mitigate disaster situations. In this article, we describe a software resource to efficiently (consuming a small run-time) locate a set of susceptible persons given a global database of user check-ins and a set of infected people. Specifically, we describe a system, named InfecTracer, that seeks to find out cases of close proximity of a person with another infected person.
As NLP models become larger, executing a trained model requires significant computational resources incurring monetary and environmental costs. To better respect a given inference budget, we propose a modification to contextual representation fine-tuning which, during inference, allows for an early (and fast) "exit" from neural network calculations for simple instances, and late (and accurate) exit for hard instances. To achieve this, we add classifiers to different layers of BERT and use their calibrated confidence scores to make early exit decisions. We test our proposed modification on five different datasets in two tasks: three text classification datasets and two natural language inference benchmarks. Our method presents a favorable speed/accuracy tradeoff in almost all cases, producing models which are up to five times faster than the state of the art, while preserving their accuracy. Our method also requires almost no additional training resources (in either time or parameters) compared to the baseline BERT model. Finally, our method alleviates the need for costly retraining of multiple models at different levels of efficiency; we allow users to control the inference speed/accuracy tradeoff using a single trained model, by setting a single variable at inference time. We publicly release our code.
Self-supervised monocular depth estimation presents a powerful method to obtain 3D scene information from single camera images, which is trainable on arbitrary image sequences without requiring depth labels, e.g., from a LiDAR sensor. In this work we present a new self-supervised semantically-guided depth estimation (SGDepth) method to deal with moving dynamic-class (DC) objects, such as moving cars and pedestrians, which violate the static-world assumptions typically made during training of such models. Specifically, we propose (i) mutually beneficial cross-domain training of (supervised) semantic segmentation and self-supervised depth estimation with task-specific network heads, (ii) a semantic masking scheme providing guidance to prevent moving DC objects from contaminating the photometric loss, and (iii) a detection method for frames with non-moving DC objects, from which the depth of DC objects can be learned. We demonstrate the performance of our method on several benchmarks, in particular on the Eigen split, where we exceed all baselines without test-time refinement.
This paper strives to localize the temporal extent of an action in a long untrimmed video. Where existing work leverages many examples with their start, their ending, and/or the class of the action during training time, we propose few-shot common action localization. The start and end of an action in a long untrimmed video is determined based on just a hand-full of trimmed video examples containing the same action, without knowing their common class label. To address this task, we introduce a new 3D convolutional network architecture able to align representations from the support videos with the relevant query video segments. The network contains: (\textit{i}) a mutual enhancement module to simultaneously complement the representation of the few trimmed support videos and the untrimmed query video; (\textit{ii}) a progressive alignment module that iteratively fuses the support videos into the query branch; and (\textit{iii}) a pairwise matching module to weigh the importance of different support videos. Evaluation of few-shot common action localization in untrimmed videos containing a single or multiple action instances demonstrates the effectiveness and general applicability of our proposal.
We propose two new Q-learning algorithms, Full-Q-Learning (FQL) and Elimination-Based Half-Q-Learning (HQL), that enjoy improved efficiency and optimality in the full-feedback and the one-sided-feedback settings over existing Q-learning algorithms. We establish that FQL incurs regret $\tilde{O}(H^2\sqrt{ T})$ and HQL incurs regret $ \tilde{O}(H^3\sqrt{ T})$, where $H$ is the length of each episode and $T$ is the total number of time periods. Our regret bounds are not affected by the possibly huge state and action space. Our numerical experiments using the classical inventory control problem as an example demonstrate the superior efficiency of FQL and HQL, and shows the potential of tailoring reinforcement learning algorithms for richer feedback models, which are prevalent in many natural problems.