Image clustering is a crucial but challenging task in machine learning and computer vision. Existing methods often ignore the combination between feature clustering. To tackle this problem, we propose Deep Adaptive Clustering (DAC) that recasts problem into binary pairwise-classification framework to judge whether pairs of images belong same clusters. In DAC, similarities are calculated as cosine distance label features which generated by deep convolutional network (ConvNet). By introducing...

Super-resolution from a single image plays an important role in many computer vision systems. However, it is still challenging task, especially preserving local edge structures. To construct high-resolution images while the sharp edges, effective edge-directed super-resolution method presented this paper. An adaptive self-interpolation algorithm first proposed to estimate gradient field directly input low-resolution image. The obtained then regarded as constraint or edge-preserving...

The present study examined the developmental origin of 'blue lies', a pervasive form lying in adult world that is told purportedly to benefit collective. Seven, 9-, and 11-year-old Chinese children were surreptitiously placed real-life situation where they decided whether lie conceal their group's cheating behavior. Children also assessed terms willingness hypothetical situations endorse or truth-telling benefits collective but at same time harms an individual. Results showed as age...

This brief presents a framework of retargeted least squares regression (ReLSR) for multicategory classification. The core idea is to directly learn the targets from data other than using traditional zero-one matrix as targets. learned target can guarantee large margin constraint requirement correct classification each point. Compared with (LSR) and recently proposed discriminative LSR models, ReLSR much more accurate in measuring error model. Furthermore, single compact model, hence there no...

Encoder-decoder models have been widely used in image captioning, and most of them are designed via single long short term memory (LSTM). The capacity single-layer network, whose encoder decoder integrated together, is limited for such a complex task captioning. Moreover, how to effectively increase the "vertical depth" encoder-decoder remains be solved. To deal with these problems, novel deep hierarchical network proposed where structure explored separate functions decoder. This model...

Multiimage super-resolution (MISR), as one of the most promising directions in remote sensing, has become a needy technique satellite market. A sequence images collected by satellites often plenty views and long time span, so integrating multiple low-resolution into high-resolution image with details emerges challenging problem. However, MISR methods based on deep learning cannot make full use images. Their fusion modules are incapable adapting to an weak temporal correlations well. To cope...

Sodium dodecyl sulfate (SDS) is widely used as a surfactant for dust suppression in coal mines to improve the collection efficiency of dust. Dust that settles may become airborne again after losing moisture through evaporation, posing an explosion risk. The extent which residual SDS affects ignition sensitivity currently unclear. This study investigates raw bituminous dust, soaked deionized water, and sodium solutions three different concentrations (0.1%, 0.3%, 0.5%). results indicate...

We propose to integrate spectral-spatial feature extraction and tensor discriminant analysis for hyperspectral image classification. First, we apply remarkable approaches in the cube extract a each pixel. Then, based on class label information, local is used remove redundant information subsequent classification procedure. The approach not only extracts sufficient features from original images but also gets better representation owing framework. Comparative results two benchmarks demonstrate...

Clustering is a crucial but challenging task in pattern analysis and machine learning. Existing methods often ignore the combination between representation learning clustering. To tackle this problem, we reconsider clustering from its definition to develop Deep Self-Evolution (DSEC) jointly learn representations cluster data. For purpose, recast as binary pairwise-classification problem estimate whether pairwise patterns are similar. Specifically, similarities defined by dot product...

Abstract BACKGROUND Optimum cultivation and management measures are needed to increase the phosphorus (P) absorption efficiency of crops for sustainable agricultural production. Previous studies indicated that leguminous can promote P by neighboring gramineous crops. In this study, we isolated screened phosphate‐solubilizing bacteria (PSB) from soybean rhizosphere under a maize–soybean intercropping system in Southwest China, nine PSBs with high P‐solubilizing ability were identified....

<title>Abstract</title> In low-light scenarios, especially under high dynamic range (HDR) conditions, existing image enhancement algorithms face significant challenges, such as over-enhancement of brightness, loss details in high-light regions, and undesirable color distortions, which severely hinder their performance. To address these limitations, we propose an improved algorithm, Retinexformer on Performer Attention (RPA). By incorporating more efficient relative position encoding, RPA...

In this paper, we present a fast image upsampling method within two-scale framework to ensure the sharp construction of upsampled for both large-scale edges and small-scale structures. our approach, low-frequency is recovered via novel sharpness preserving interpolation technique based on well-constructed displacement field, which estimated by cross-resolution model. Within model, distances pixels are preserved, enables recovery in high-resolution result. Likewise, local high-frequency...

Due to the factors such as visual occlusion, illumination change and pose variation, it is a challenging task develop effective efficient models for vehicle detection classification in surveillance videos. Although plenty of existing related have been proposed, many issues still need be resolved. Typically, methods should vulnerable complex environments. Moreover, spite thoughtful attempts on adaptive appearance solve occlusion problem, corresponding approaches often suffer from high...

Traditional clustering methods often perform with low-level indiscriminative representations and ignore relationships between patterns, resulting in slight achievements the era of deep learning. To handle this problem, we develop Deep Discriminative Clustering (DDC) that models task by investigating patterns a neural network. Technically, global constraint is introduced to adaptively estimate relationships, local developed endow network capability learning high-level discriminative...

Deep learning has achieved huge success in the field of artificial intelligence, but performance heavily depends on labeled data. Few-shot aims to make a model rapidly adapt unseen classes with few samples after training base dataset, and this is useful for tasks lacking data such as medical image processing. Considering that core problem few-shot lack samples, straightforward solution issue augmentation. This paper proposes generative (VI-Net) based cosine-classifier baseline. Specifically,...

The detection of virus RNA in wastewater has been established as a valuable method for monitoring Coronavirus disease 2019. Carbon nanomaterials hold potential application separating owing to their effective adsorption and extraction capabilities. However, carbon have limited separability under homogeneous aqueous conditions. Due the stabilities nanostructure, it is challenge efficiently immobilize them onto magnetic beads separation. Here, we develop porous agarose layered graphene oxide...

The image manipulation detection localization task differs from traditional computer vision tasks in that we focus more on capturing subtle and generic features images. In this paper, propose a novel method called irrelevant visual information suppression, which aims to alleviate the interference of images feature extraction, thereby obtaining traces are unrelated semantic information. general, most operations leave at edges. Therefore, introduce specially designed manipulated edge...