With the development of deep learning, the benchmark of hyperspectral imagery classification is constantly improving, but there are still significant challenges for hyperspectral imagery classification of few-shot scenes. This letter proposes an active-learning-based prototypical network (ALPN), which uses the prototypical network to extract representative features from a few samples. Moreover, it combines semisupervised clustering and active learning methods to select and request labels from valuable examples actively. In this way, the feature extraction ability of the network is gradually optimized. The experimental results validated that the classification accuracy and robustness of ALPN significant exceeded the comparison baselines. Furthermore, because it can be regarded as a sample selection method, ALPN can be easily combined with other models to obtain better classification results.

With the development of deep learning, hyperspectral image classification (HSIC) has improved rapidly in recent years. Unsupervised feature learning algorithms play an important role in extracting features from hyperspectral images (HSIs). This letter proposed a random-occlusion-based Bootstrap-Your-Own-Latent network (ROBYOL), combining a new augmentation method and a superior contrastive learning algorithm for feature extraction. The proposed method consists of a self-supervised learning part for feature extracting and a classifier part as the downstream task. It can be proved by the experimental results that the feature extraction ability of the network is effective in this way. Furthermore, the influence of different occlusion strategies is also studied, including changing occlusion area and occlusion value, and we proposed translucent occlusion. Our results with two well-known HSIs reveal that proper occlusion strategies can improve hyperspectral classification results effectively.

Improving the accuracy of the forecasting of building power consumption is helpful in reducing commercial expenses and carbon emissions. However, challenges such as the shortage of training data and the absence of efficient models are the main obstacles in this field. To address these issues, this work introduces a model named MTLTFT, combining multi-task learning (MTL) with the temporal fusion transformer (TFT). The MTL approach is utilized to maximize the effectiveness of the limited data by introducing multiple related forecasting tasks. This method enhances the learning process by enabling the model to learn shared representations across different tasks, although the physical number of data remains unchanged. The TFT component, which is optimized for feature learning, is integrated to further improve the model’s performance. Based on a dataset from a large exposition building in Hangzhou, we conducted several forecasting experiments. The results demonstrate that MTLTFT outperforms most baseline methods (such as LSTM, GRU, N-HiTS) in terms of Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE), suggesting that MTLTFT is a promising approach for the forecasting of building power consumption and other similar tasks.

Predicting energy consumption in large exposition centers presents a significant challenge, primarily due to the limited datasets and fluctuating electricity usage patterns. This study introduces a cutting-edge algorithm, the contrastive transformer network (CTN), to address these issues. By leveraging self-supervised learning, the CTN employs contrastive learning techniques across both temporal and contextual dimensions. Its transformer-based architecture, tailored for efficient feature extraction, allows the CTN to excel in predicting energy consumption in expansive structures, especially when data samples are scarce. Rigorous experiments on a proprietary dataset underscore the potency of the CTN in this domain.

The noise corruption problem commonly exists in hyperspectral images (HSIs) and severely affects the accuracy of hyperspectral unmixing algorithms. The noise formulation existing in HSIs is relatively complex and would change in conjunction with different devices and imaging settings. For real applications, applying denoising approaches without accurate close-to-reality noise modeling before unmixing may not improve, but rather degrade the unmixing performance. This study proposes a robust hyperspectral unmixing method with practical learning-based hyperspectral image denoising. We formulated a close-to-reality noise model for hyperspectral data and provide a calibration approach for the noise parameters. On the basis of the calibrated noise model, synthetic data were generated and used for training a KST-based denoising network. The noisy hyperspectral data were firstly denoised by the trained denoising network and were then used to perform the unmixing process. A variety of unmixing algorithms can be integrated into our method to improve the accuracy of unmixing in noisy situations. In the experiments, several widely used unmixing algorithms were employed to verify the effect of the proposed method. The experimental results on both synthetic and real demonstrated that our proposed method can handle HSI data with various gain settings and helps to improve the unmixing performance effectively.