Identifying multifunctional therapeutic peptides (MFTP) is an important yet complex challenge in the realm of peptide recognition. Unlike monofunctional peptides, MFTP classification requires discerning fine-grained labeling information associated with amino acids, making it more intricate. Existing methods often ignore the nuanced semantics of these labels and fail to fully explore the interplay between peptide sequences and their labels. To address these issues, we propose a multilabel classification method named MultiPep-DLCL. This method uses a deep learning-based model architecture to translate peptide sequences into sequence features by learning the local and global dependencies of multifunctional therapeutic peptide sequences. Additionally, the Label-Sequence Fusion Transformer is employed to efficiently learn high-quality label embeddings by mining effective information from peptide sequences. Finally, the correspondence between sequence features and label embeddings is strengthened through label-sequence contrastive learning. To tackle dataset imbalance, MultiPep-DLCL integrates a multilabel focal dice loss function alongside the traditional cross-entropy loss function. Experimental results demonstrate that the MultiPep-DLCL significantly outperforms existing methods in MFTP recognition.

Li et al. in [Inf. Process. Lett. 77 (2001) 35–41] proposed the shuffle-cube$SQ_{n}$, a hypercube variant, as an attractive interconnection network topology for massive parallel and distributed systems. Diameter and symmetry are two desirable measures of network performance in terms of transmission delay and routing algorithms. Almost all$n$-regular hypercube variants of dimension$n$have diameter not less than$n/2$. The diameter of the shuffle-cube is approximately a quarter of the diameter of the hypercube of the same dimension, making it a competitive candidate network topology. By far, symmetric properties of the shuffle-cube remain unknown. In this paper, we show that$SQ_{n}$is not vertex-transitive for$n\gt 2$, which is not an appealing property in interconnection networks. This shortcoming limits the practical application of the shuffle-cube. To overcome this limitation, two novel variants of the shuffle-cube, namely simplified shuffle-cube$SSQ_{n}$and balanced shuffle-cube$BSQ_{n}$are introduced, and their vertex-transitivity are proved simultaneously. By proposing the shuffle-cube-like graph, we obtain that both$SSQ_{n}$and$BSQ_{n}$are maximally connected, implying high connectivity similar to the hypercube. Additionally, super-connectivity, a refined parameter of connectivity, of$SSQ_{n}$and$BSQ_{n}$are also determined. Then, by vertex-transitivity of$SSQ_{n}$and$BSQ_{n}$, routing algorithms of$SSQ_{n}$and$BSQ_{n}$are given for all$n\gt 2$respectively. We show that both$SSQ_{n}$and$BSQ_{n}$possess Hamiltonian cycle embedding for all$n\gt 2$, and we also show that$SSQ_{n}$is Hamiltonian-connected. It is noticeable that each vertex of$SSQ_{n}$is contained in exactly one clique of size four, making it also a viable interconnection topology for data center networking since each clique of size four can be viewed as an efficient local data processing cluster of the network. Finally, as a by-product of proving vertex-transitivity of$BSQ_{n}$, we mend a flaw in the Property 3 in [IEEE Trans. Comput. 46 (1997) 484–490].

The primary challenge in image segmentation arises from the scale selection issue induced by the complex structure of scenes portrayed as an image. To solve this problem, this study developed a novel framework named scene-specific unsupervised segmentation parameter optimization to conduct multiscale segmentation. Images were firstly divided into different partitions based on the complexity of scene structures, where the structural complexity was measured using an extended structural similarity index. Subsequently, parameter optimization was implemented separately within each image partition. The results indicated significant variations in the estimated optimal segmentation scales across different partitions, following the corresponding scene structural complexity within each partition. Moreover, the accuracy evaluation results demonstrated the effectiveness of our methodology in segmenting geographic objects with diverse geometric and spectral features, significantly improving the overall segmentation quality.

To efficiently reconstruct magnetic resonance images (MRI) from highly undersampled measurements by using compressed sensing (CS), in this letter, we propose a hybrid regularization model from deep prior and low-rank prior. The local deep prior is explored by a fast flexible denoising convolutional neural network (FFDNet). To compensate for 1) the generalization capability of FFDNet on artifact noise caused by undersampling K-space and 2) the inaccurate noise estimation for various undersampling ratios, we model the low-rank prior as a weighted Schatten p-norm to obtain the global information of MRIs. The final model, combined by the local deep and low-rank priors, is solved by the alternating directional method of multipliers under the plug-and-play framework. Compared with the popular CS-MRI approaches, the experimental results demonstrate that the proposed method can achieve better reconstruction performance in terms of quality index and visual effects.

Quantization aims to form new vectors or matrices with shared values close to the original. In recent years, the popularity of scalar quantization has been soaring as it is found huge utilities in reducing the resource cost of neural networks. Popular clustering-based techniques suffers substantially from the problems of dependency on the seed, empty or out-of-the-range clusters, and high time complexity. To overcome the problems, in this paper, scalar quantization is examined from a new perspective, namely sparse least square optimization. Specifically, several quantization algorithms based on l1l1 least square are proposed and implemented. In addition, similar schemes with l1+ l2l1+l2 and l0l0 regularization are proposed. Furthermore, to compute quantization results with given amount of values/clusters, this paper proposes an iterative method and a clustering-based method, and both of them are built on sparse least square optimization. The algorithms proposed are tested under three data scenarios and their computational performance, including information loss, time consumption, and distribution of values of sparse vectors are compared. The paper offers a new perspective to probe the area of quantization, and the algorithms proposed are superior especially under bit-width reduction scenarios, where the required post-quantization resolution (the number of values) is not significantly lower than the original scalar.

In the imaging process of the remote sensing, there was degradation phenomenon in the acquired images. This paper, we present a group sparse regularization based iterative incremental for remote image deblurring estimating a single latent sharp image given either a single or multiple blurry and/or noisy observations. Considering the sparseness and nonlocal similarity properties of image, a group sparse representation based incremental iterative method is established for blurry image restoration. The proposed sparse representation enforces the intrinsic local sparsity and nonlocal self-similarity of images simultaneously in a unified framework. Comprehensive experiments demonstrate that the framework integrating the sparseness property of images significantly improves the deblurring performance.

As manually correcting error in visual interpretation data is both costly and labor intensive, the automated error detection and correction approaches is in need. In this paper, a prototype method is developed to detect and correct errors in visually interpreted landuse data. This method involves image segmentation, anomaly detection and decision tree classification techniques. The method is tested on landuse dataset with known accuracy (95%-50%). Result shows that the accuracy of landuse data can be greatly improved by the error correction method, and the approach can be practical when the accuracy of the interpretation data is no less than 70%.

In this paper, the authors had a research on the winter wheat yield estimation using retrieved LAI from remote sensing in typical 11 counties in Huanghuaihai Plain in North China. In order to improve the quality of data and reduce error of yield estimation, Savitzky-Golay filter (S-G filter) was used to smooth the NDVI series to reduce influences of cloud contamination and abnormal data. At the same time, Gaussian model was used to simulate daily crop LAI to get average LAI at each growth stage. Using these average LAI, the authors established relationships between LAI and yield of winter wheat at main growth stage. After optimization of yield estimation model, the best period of time and best model was selected out. Finally, the authors depended on retrieved LAI from MODIS-NDVI to estimate winter wheat yield. The results showed that average relative error was 1.21% and that RMSE was 257.33 kg ha−1comparing predicted yield with ground truth data. We draw a conclusion that we could accurately get winter wheat yield about 20–30 days ahead of harvest time.