Gesture recognition aims to recognize meaningful movements of human bodies, and is of utmost importance in intelligent human-computer/robot interactions. In this paper, we present a multimodal gesture recognition method based on 3-D convolution and convolutional long-short-term-memory (LSTM) networks. The proposed method first learns short-term spatiotemporal features of gestures through the 3-D convolutional neural network, and then learns long-term spatiotemporal features by convolutional LSTM networks based on the extracted short-term spatiotemporal features. In addition, fine-tuning among multimodal data is evaluated, and we find that it can be considered as an optional skill to prevent overfitting when no pre-trained models exist. The proposed method is verified on the ChaLearn LAP large-scale isolated gesture data set (IsoGD) and the Sheffield Kinect gesture (SKIG) data set. The results show that our proposed method can obtain the state-of-the-art recognition accuracy (51.02% on the validation set of IsoGD and 98.89% on SKIG).

Single-cell RNA-sequencing (scRNA-seq) is a recent breakthrough technology, which paves the way for measuring RNA levels at single cell resolution to study precise biological functions. One of the main challenges when analyzing scRNA-seq data is the presence of zeros or dropout events, which may mislead downstream analyses. To compensate the dropout effect, several methods have been developed to impute gene expression since the first Bayesian-based method being proposed in 2016. However, these methods have shown very diverse characteristics in terms of model hypothesis and imputation performance. Thus, largescale comparison and evaluation of these methods is urgently needed now. To this end, we compared eight imputation methods, evaluated their power in recovering original real data, and performed broad analyses to explore their effects on clustering cell types, detecting differentially expressed genes, and reconstructing lineage trajectories in the context of both simulated and real data. Simulated datasets and case studies highlight that there are no one method performs the best in all the situations. Some defects of these methods such as scalability, robustness and unavailability in some situations need to be addressed in future studies.

Alzheimer's disease (AD) is an irreversible brain disease that severely damages human thinking and memory. Early diagnosis plays an important part in the prevention and treatment of AD. Neuroimaging-based computer-aided diagnosis (CAD) has shown that deep learning methods using multimodal images are beneficial to guide AD detection. In recent years, many methods based on multimodal feature learning have been proposed to extract and fuse latent representation information from different neuroimaging modalities including magnetic resonance imaging (MRI) and 18-fluorodeoxyglucose positron emission tomography (FDG-PET). However, these methods lack the interpretability required to clearly explain the specific meaning of the extracted information. To make the multimodal fusion process more persuasive, we propose an image fusion method to aid AD diagnosis. Specifically, we fuse the gray matter (GM) tissue area of brain MRI and FDG-PET images by registration and mask coding to obtain a new fused modality called “GM-PET.” The resulting single composite image emphasizes the GM area that is critical for AD diagnosis, while retaining both the contour and metabolic characteristics of the subject's brain tissue. In addition, we use the three-dimensional simple convolutional neural network (3D Simple CNN) and 3D Multi-Scale CNN to evaluate the effectiveness of our image fusion method in binary classification and multi-classification tasks. Experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset indicate that the proposed image fusion method achieves better overall performance than unimodal and feature fusion methods, and that it outperforms state-of-the-art methods for AD diagnosis.

Continuous gesture recognition aims at recognizing the ongoing gestures from continuous gesture sequences, and is more meaningful for the scenarios where the start and end frames of each gesture instance are generally unknown in practical applications. This paper presents an effective deep architecture for continuous gesture recognition. Firstly, continuous gesture sequences are segmented into isolated gesture instances using the proposed temporal dilated Res3D network. A balanced squared hinge loss function is proposed to deal with the imbalance between boundaries and non-boundaries. Temporal dilation can preserve the temporal information for the dense detection of the boundaries at fine granularity, and the large temporal receptive field makes the segmentation results more reasonable and effective. Then, the recognition network is constructed based on the 3D convolutional neural network (3DCNN), the convolutional Long-Short-Term-Memory network (ConvLSTM), and the 2D convolutional neural network (2DCNN) for isolated gesture recognition. The "3DCNN-ConvLSTM-2DCNN" architecture is more effective to learn long-term and deep spatiotemporal features. The proposed segmentation and recognition networks obtain the Jaccard index of 0.7163 on the Chalearn LAP ConGD dataset, which is 0.106 higher than the winner of 2017 ChaLearn LAP Large-scale Continuous Gesture Recognition Challenge.

Convolutional long short-term memory (ConvLSTM) networks have been widely used for action/gesture recognition, and different attention mechanisms have also been embedded into ConvLSTM networks. This paper explores the redundancy of spatial convolutions and the effects of the attention mechanism in ConvLSTM, based on our previous gesture recognition architectures that combine the 3-D convolutional neural network (CNN) and ConvLSTM. Depthwise separable, group, and shuffle convolutions are used to replace the convolutional structures in ConvLSTM for the redundancy analysis. In addition, four ConvLSTM variants are derived for attention analysis: 1) by removing the convolutional structures of the three gates in ConvLSTM; 2) by applying the attention mechanism on the ConvLSTM input; and 3) by reconstructing the input and 4) output gates with the modified channelwise attention mechanism. Evaluation results demonstrate that the spatial convolutions in the three gates scarcely contribute to the spatiotemporal feature fusion and that the attention mechanisms embedded into the input and output gates cannot improve the feature fusion. In other words, ConvLSTM mainly contributes to the temporal fusion along with the recurrent steps to learn long-term spatiotemporal features when taking spatial or spatiotemporal features as input. A new LSTM variant is derived on this basis in which the convolutional structures are embedded only into the input-to-state transition of LSTM. The code of the LSTM variants is publicly available.\footnotehttps://github.com/GuangmingZhu/ConvLSTMForGR.

The broad learning system (BLS) is designed based on the technology of compressed sensing and pseudo-inverse theory, and consists of feature nodes and enhancement nodes, has been proposed recently. Compared with the popular deep learning structures, such as deep neural networks, BLS has the ability of rapid incremental learning and can remodel the system without the usual tedious retraining process. However, given that BLS is still in its infancy, it still needs analysis, improvements, and verification. In this article, we first analyze the principle of fast incremental learning ability of BLS in depth. Second, in order to provide an in-depth analysis of the BLS structure, according to the novel structure design concept of deep neural networks, we present four brand-new BLS variant networks and their incremental realizations. Third, based on our analysis of the effect of feature nodes and enhancement nodes, a new BLS structure with a semantic feature extraction layer has been proposed, which is called SFEBLS. The experimental results show that SFEBLS and its variants can increase the accuracy rate on the NORB dataset 6.18%, Fashion-MNIST dataset by 3.15%, ORL data by 5.00%, street view house number dataset by 12.88%, and CIFAR-10 dataset by 18.42%, respectively, and the four brand-new BLS variant networks also obviously outperform the original BLS.