Classification of hyperspectral image (HSI) is an important research topic in the remote sensing community. Significant efforts (e.g., deep learning) have been concentrated on this task. However, it is still an open issue to classify the high-dimensional HSI with a limited number of training samples. In this paper, we propose a semi-supervised HSI classification method inspired by the generative adversarial networks (GANs). Unlike the supervised methods, the proposed HSI classification method is semi-supervised, which can make full use of the limited labeled samples as well as the sufficient unlabeled samples. Core ideas of the proposed method are twofold. First, the three-dimensional bilateral filter (3DBF) is adopted to extract the spectral-spatial features by naturally treating the HSI as a volumetric dataset. The spatial information is integrated into the extracted features by 3DBF, which is propitious to the subsequent classification step. Second, GANs are trained on the spectral-spatial features for semi-supervised learning. A GAN contains two neural networks (i.e., generator and discriminator) trained in opposition to one another. The semi-supervised learning is achieved by adding samples from the generator to the features and increasing the dimension of the classifier output. Experimental results obtained on three benchmark HSI datasets have confirmed the effectiveness of the proposed method, especially with a limited number of labeled samples.

Reinforcement learning (RL)-based brain machine interfaces (BMIs) enable the user to learn from the environment through interactions to complete the task without desired signals, which is promising for clinical applications. Previous studies exploited Q-learning techniques to discriminate neural states into simple directional actions providing the trial initial timing. However, the movements in BMI applications can be quite complicated, and the action timing explicitly shows the intention when to move. The rich actions and the corresponding neural states form a large state-action space, imposing generalization difficulty on Q-learning. In this paper, we propose to adopt attention-gated reinforcement learning (AGREL) as a new learning scheme for BMIs to adaptively decode high-dimensional neural activities into seven distinct movements (directional moves, holdings and resting) due to the efficient weight-updating. We apply AGREL on neural data recorded from M1 of a monkey to directly predict a seven-action set in a time sequence to reconstruct the trajectory of a center-out task. Compared to Q-learning techniques, AGREL could improve the target acquisition rate to 90.16% in average with faster convergence and more stability to follow neural activity over multiple days, indicating the potential to achieve better online decoding performance for more complicated BMI tasks.

No abstract

Brain-machine interfaces (BMIs) translate neural activities of the brain into specific instructions that can be carried out by external devices. BMIs have the potential to restore or augment motor functions of paralyzed patients suffering from spinal cord damage. The neural activities have been used to predict the 2D or 3D movement trajectory of monkey's arm or hand in many studies. However, there are few studies on decoding the wrist movement from neural activities in center-out paradigm. The present study developed an invasive BMI system with a monkey model using a 10×10-microelectrode array in the primary motor cortex. The monkey was trained to perform a two-dimensional forelimb wrist movement paradigm where neural activities and movement signals were simultaneous recorded. Results showed that neuronal firing rates highly correlated with forelimb wrist movement; > 70% (105/149) neurons exhibited specific firing changes during movement and > 36% (54/149) neurons were used to discriminate directional pairs. The neuronal firing rates were also used to predict the wrist moving directions and continuous trajectories of the forelimb wrist. The four directions could be classified with 96% accuracy using a support vector machine, and the correlation coefficients of trajectory prediction using a general regression neural network were above 0.8 for both horizontal and vertical directions. Results showed that this BMI system could predict monkey wrist movements in high accuracy through the use of neuronal firing information.

No abstract