Object Of Interest Research Articles

Bridging the gap between object detection in close-up and high-resolution wide shots

Recent years have seen a significant rise in gigapixel-level image/video capture systems and benchmarks with high-resolution wide (HRW) shots. Different from close-up shots like MS COCO, the higher resolution and wider field of view raise new research and application problems, such as how to perform accurate and efficient object detection with such large input in low-power edge devices like UAVs. There are several unique challenges in HRW shots. (1) Sparse information: the objects of interest cover less area. (2) Various scale: there is 10 to 100× object scale change in one single image. (3) Incomplete objects: the sliding window strategy to handle the large input leads to truncated objects at the window edge. (4) Multi-scale information: it is unclear how to use multi-scale information in training and inference. Consequently, directly using a close-up detector leads to inaccuracy and inefficiency. In this paper, we systematically investigate this problem and bridge the gap between object detection in close-up and HRW shots, by introducing a novel sparse architecture that can be integrated with common networks like ConvNet and Transformer. It leverages alternative sparse learning to complementarily fuse coarse-grained and fine-grained features to (1) adaptively extract valuable information from (2) different object scales. We also propose a novel Cross-window Non-Maximum Suppression (C-NMS) algorithm to (3) improve the box merge from different windows. Furthermore, we propose a (4) simple yet effective multi-scale training and inference strategy to improve accuracy. Experiments on two benchmarks with HRW shots, PANDA and DOTA-v1.0, demonstrate that our methods significantly improve accuracy (up to 5.8%) and speed (up to 3×) over SotAs, for both ConvNet or Transformer based detectors, on edge devices. Our code is open-sourced and available at https://github.com/liwenxi/SparseFormer.

Latent Space Bayesian Optimization With Latent Data Augmentation for Enhanced Exploration.

Latent space Bayesian optimization (LSBO) combines generative models, typically variational autoencoders (VAE), with Bayesian optimization (BO), to generate de novo objects of interest. However, LSBO faces challenges due to the mismatch between the objectives of BO and VAE, resulting in poor exploration capabilities. In this article, we propose novel contributions to enhance LSBO efficiency and overcome this challenge. We first introduce the concept of latent consistency/inconsistency as a crucial problem in LSBO, arising from the VAE-BO mismatch. To address this, we propose the latent consistent aware-acquisition function (LCA-AF) that leverages consistent points in LSBO. Additionally, we present LCA-VAE, a novel VAE method that creates a latent space with increased consistent points through data augmentation in latent space and penalization of latent inconsistencies. Combining LCA-VAE and LCA-AF, we develop LCA-LSBO. Our approach achieves high sample efficiency and effective exploration, emphasizing the significance of addressing latent consistency through the novel incorporation of data augmentation in latent space within LCA-VAE in LSBO. We showcase the performance of our proposal via de novo image generation and de novo chemical design tasks.

Algorithm for constructing unmeasured logging curves (pseudo-logging) in wells of developed fields

In many cases, the diagrams of some logs in an old well of producing fields are not recorded for some reason or another. Currently, the importance of reassessing the efficiency of such fields is obvious, while the absence of some log curves causes great difficulties. In particular very few studies with acoustic logging (AL) have been carried out in old wells. On the other hand, in 1960-80, among the studies carried out in wells, preference was given to various types of electrical logging (EL), neutron-gamma logging (NGL), and gamma ray logging (GRL). To overcome such difficulties, it is necessary to create diagrams of unmeasured well logs or pseudo-logs. Nowadays, there are methods for constructing pseudolog curves. For example, the method of statistical analysis is widely used. However, in most cases, their implementation is very complex and requires a lot of work. We have developed a new method to construct pseudo-log curves. At first, the propagation velocity of elastic waves in rocks is studied based on seismic survey (SS) and AL data, which are performed presently on the investigation field. Based on this information, objects of interest to the oil industry are selected, their geometric dimensions are determined, the contour of oil and gas saturation is assessed, areas with abnormally high reservoir pressure are studied, the process of accumulation deposition is re-examined, and the preparation of an optimal well network diagram for the effective exploitation of the field is determined. Further, based on the long-ago measured curves of electrical loggings and neutron-gamma logging, we obtain the pseudo-log AL curve in old boreholes and the data SS. The proposed method was tested in 48 wells in the Bahar field (Azerbaijan). The method has shown high efficiency, its reliability is confirmed by independent data. The results of the method help in solving several geological problems: monitoring changes in petrophysical quantities of rocks between wells during development; study of well formations located below the design depth; recalculation of hydrocarbon reserves; determination of the location for drilling new wells, etc

How Do Observational Astronomers Learn to Inspect Imaging Data

Astronomy is entering an unprecedented era of data collection. Upcoming large surveys will gather more data than ever before, generated at rates requiring real-time decision making. Looking ahead, it is inevitable that astronomers will need to rely more heavily on automated processes. Indeed, some instances have already arisen wherein the majority of the inspection process is automated. Visual discovery, performed traditionally by humans, is one key area where automation is now being integrated rapidly. Visual discovery comprises two aspects: (1) visual inspection, the skill associated with examining an image to identify areas or objects of interest; and (2) visual interpretation, the knowledge associated with the classification of the objects or features. Both skills and knowledge are vital for humans to perform visual discovery, however, there appears to have been limited investigation into how the skill of visual inspection in astronomy is acquired. In this work, we address this issue by setting out to identify the landscape within which observational astronomers develop the skills to perform visual inspection. We report on a survey of 70 professional observational astronomers, at various career stages and from different geographical regions. We found that between 63%and 73% of the astronomers surveyed had received formal and/or informal training in visual inspection of images, although formal training (21%) was less common than informal training (60%). Surprisingly, out of the 37% who did not recall having received training in visual inspection, 29% (20 astronomers) indicated that they provided training to others. This suggests the emergence of ``expertise without precedent'' where a first expert in the field provides a new way of achieving a task. These results, paired with a set of three pilot interviews, present a touchstone against which the training of future observational astronomers can be compared.

Sequence Segmentation of Nematodes in Atlantic Cod with Multispectral Imaging Data.

Nematodes pose significant challenges for the fish processing industry, particularly in white fish. Despite technological advances, the industry still depends on manual labor for the detection and extraction of nematodes. This study addresses the initial steps of automatic nematode detection and differentiation from other common defects in fish fillets, such as skin remnants and blood spots. VideometerLab 4, an advanced Multispectral Imaging (MSI) System, was used to acquire 270 images of 50 Atlantic cod fillets under controlled conditions. In total, 173 nematodes were labeled using the Segment Anything Model (SAM), which is trained to automatically segment objects of interest from only few representative pixels. With the acquired dataset, we study the potential of identifying nematodes through their spectral signature. We incorporated normalized Canonical Discriminant Analysis (nCDA) to develop segmentation models trained to distinguish between different components within the fish fillets. By incorporating multiple segmentation models, we aimed to achieve a satisfactory balance between false negatives and false positives. This resulted in 88% precision and 79% recall for our annotated test data. This approach could improve process control by accurately identifying fillets with nematodes. Using MSI minimizes unnecessary inspection of fillets in good condition and concurrently boosts product safety and quality.

NON-INVASIVE FORENSIC EXAMINATION OF INSCRIPTIONS IN A HISTORICAL DOCUMENT AS AN ATTEMPT TO EVALUATE THE DIVERSITY OF THE COMMERCIAL OFFER OF INKS

In the history of the German state, the world-wide economic crisis (1929-1939) probably also had an impact on the availability of inks. The research aimed to find out whether it is possible to discriminate between inks and determine the prevalence of ink manufacturers in a specific area, which, in statistical terms, could provide insight into the development of regional ink producers. The research covered selected inscriptions in ink written in a Diary from a small geographical area. Ink examination was carried out using non-invasive VIS-NIR and RS spectroscopic methods involving a statistical approach. The obtained results allowed effective discrimination of the inks and divided them into several groups. Research has shown that iron-gall inks from several ink manufacturers were used to prepare the entries in the Diary. It cannot be ruled out that the differences in individual inks resulted from technological changes in their production process. Additionally, the research revealed a multitude of entries made with iron-gall inks, which may testify to the development of small-scale production of iron-gall inks confined to one geographical area. This type of historical document may constitute a forgery in whole or in specific parts of the inscriptions executed to falsify significant facts, therefore it is an object of interest of forensic science. The non-invasive VIS-NIR and RS spectroscopic methods supported by statistical approach should continue to develop to become an alternative and reliable solution offering great potential in forensic examination of inks.

Zero-shot image segmentation for monitoring thermal conditions of individual cage-free laying hens

Body temperature is a critical indicator of the health and productivity of egg-laying chickens and other domesticated animals. Recent advancements in thermography allow for precise surface temperature measurement without physical contact with animals, reducing animal stress from human handling. Gold standard temperature analysis via thermography requires manual selection of limited points for an object of interest, which could be time-consuming and inadequate for representing the comprehensive thermal profile of a chicken’s body. The objective of this study was to leverage and optimize a zero-shot artificial intelligence technology for the automatic segmentation of individual cage-free laying hens within thermal images, providing insights into their overall thermal conditions. A zero-shot image segmentation model (Segment Anything, “SAM”) was modified by replacing manual selections of target points with automatic selection of the initial point using pre-processing techniques (e.g., thresholding) in each thermal image. The model was also incorporated with post-processing techniques integrated with a machine learning classifier to improve segmentation accuracy. Three versions of modified SAM models (i.e., SAM, FastSAM, and MobileSAM), two common instance segmentation algorithms (i.e., YOLOv8 and Mask R-CNN), and two foundation segmentation models (i.e., U2-Net and ISNet) were comparatively evaluated to determine the optimal one for bird segmentation from thermal images. A total of 1,917 thermal images were collected from cage-free laying hens (Hy-Line W-36) at 77–80 weeks of age. The image dataset exhibited considerable variations such as feathers, bird movement, body gestures, and the specific conditions of cage-free facilities. The experimental results demonstrate that the modified SAM did not only surpass the six zero-shot models—YOLOv8, Mask R-CNN, FastSAM, MobileSAM, U2Net, and ISNet—but also outperformed other modified SAM-based models (Modified FastSAM and Modified MobileSAM) in terms of hen detection performance, achieving a success rate of 84.4 %, and in segmentation performance, with an intersection over union of 85.5 %, recall of 91.0 %, and an F1 score of 92.3 %. The optimal model, modified SAM, was pipelined to extract statistics including the averages (°C) of mean (27.03, 27.04, 28.53, 26.68), median (26.27, 26.84, 28.28, 26.78), 25th percentile (25.33, 25.61, 27.26, 25.53), and 75th percentile (28.04, 27.95, 29.22, 27.55) of surface body temperature of individual laying hens in thermal images for each week. More statistics of hen body surface temperature can be extracted based on the segmentation results. The developed pipeline is a useful tool for automatically evaluating the thermal conditions of individual birds.

Noninvasive Multitechnique Analysis of Baroque Amber Artworks From Gdansk: The Benefits of Using Raman Spectroscopy

ABSTRACTThe multidisciplinary, noninvasive analysis of baroque amber artworks is part of a research project that deals with the most valuable objects from the Museum of Gdansk (Poland). The two most interesting objects will be presented here: a Baroque Gdansk wardrobe‐shaped amber cabinet (made by Johann Georg Zernebach, Gdansk, 1724) and a crucifix (Gdansk, 17th century). Macro‐ (UV, IR, and X‐radiography) and micro‐ (XRF and Raman spectroscopy) analytical methods were applied to reveal traces of old conservation treatments and uncover the techniques of the 17th and 18th century amber masters. UV photography shows the differences between the amber plates that are not so easily detected under visible light, while XRF spectrometry detects elements atypical for amber objects, suggesting previous conservation treatments. Confocal Raman measurements, especially Raman depth profiling, were performed for the amber artworks and allowed to detect areas of previous conservation treatments. Single‐point scans were collected from the surface (0 μm) to a depth of −200 μm, with a step of 50 μm. The results allowed us to identify three different types of zones: where the amber was preserved without any protective layer, places where the amber was covered with a thin layer of a protective substance, and places where amber elements were compensated for loss using a binding agent and filler. The presented project allowed the development of a comprehensive methodology for the analysis of amber objects, especially to optimize the capabilities of confocal Raman microscopy. The combination of macro‐ and microanalytical techniques made it possible to obtain a broad overview of such complex artworks while optimizing the time and effort spent on the investigations.

Exoplanet Detection Using Machine Learning : A Comparative Study Using Kepler Mission Data

The search for habitable planets outside our solar system has captivated scientists throughout the centuries. Discovery and characterization of exoplanets have been one of the most important endeavors of modern astronomy. With various space missions, we have significantly expanded our observational capacity, resulting in an abundance of information about the universe. The influx of more data necessitates the development of techniques that can aid astronomers in processing all the information more efficiently and in an automated manner. Machine learning in recent years has become an indispensable paradigm to automate complex tasks that are possible only by humans. This work explores the application of machine learning to detect exoplanets from NASA’s Kepler mission. Our dataset comprises Kepler Objects of Interest (KOIs), encompassing their characteristic features and confirmed exoplanet status. We experiment with multiple supervised classification techniques including classical, tree-based, and neural methods. The best-performing model Histogram Gradient Boosting achieves a strong performance of 94.6% precision and 94.1% recall on a held-out dataset demonstrating the strong potential of integrating machine learning techniques into astronomy, potentially leading to new insights into planetary systems outside the solar system.

Rapid Efficacy of riSankizumab in pretibial psoriasis invOLVEment: RESOLVE.

Despite extraordinary improvements in the management of psoriasis in recent times, some areas of the body, such as the pretibial area, still show an unsatisfactory response and a more significant impact on patient quality of life. This multicentre study focuses on psoriasis affecting sensitive areas (particularly the pretibial area), its impact on quality of life and the therapeutic response to risankizumab. This multicentre prospective observational study recruited patients with moderate-to-severe psoriasis with pretibial area involvement. All patients underwent treatment with risankizumab (150 mg every 3 weeks), and efficacy was assessed after 24 weeks. The study included 128 patients with a mean age of 51 years, suffering from moderate-to-severe psoriasis with involvement of the pretibial area with median total Psoriasis Area Severity Index score of 17.05 and Dermatology Life Quality Index of 16.27. The group was further divided into two sub-groups: the 'mother patch' group, in whom the very first psoriatic plaque appeared in the pretibial region (45 patients), and the 'non-mother patch' group, in whom the psoriatic lesion in the pretibial region was present but not as the first manifestation (83 patients). In order to better assess the involvement of psoriasis in the pretibial area, the pretibial plaque lesion severity index was also calculated at baseline in all patients: extent 2.75, erythema 2.64, infiltration 2.45 and desquamation 2.38. All participants in this study showed a good therapeutic response, with a reduction in all scores. The pretibial area is becoming an object of therapeutic interest due to some resistance to clearance and the consequent impairment of patient quality of life. This study showed that risankizumab can give favourable therapeutic results not only in patients with moderate-to-severe psoriasis with involvement of the difficult-to-treat areas but particularly in patients with recalcitrant plaques in the pretibial area.

Hyperspectral Image Denoising and Compression Using Optimized Bidirectional Gated Recurrent Unit

The availability of a higher resolution fine spectral bandwidth in hyperspectral images (HSI) makes it easier to identify objects of interest in them. The inclusion of noise into the resulting collection of images is a limitation of HSI and has an adverse effect on post-processing and data interpretation. Denoising HSI data is thus necessary for the effective execution of post-processing activities like image categorization and spectral unmixing. Most of the existing models cannot handle many forms of noise simultaneously. When it comes to compression, available compression models face the problems of increased processing time and lower accuracy. To overcome the existing limitations, an image denoising model using an adaptive fusion network is proposed. The denoised output is then processed through a compression model which uses an optimized deep learning technique called "chaotic Chebyshev artificial hummingbird optimization algorithm-based bidirectional gated recurrent unit" (CCAO-BiGRU). All the proposed models were tested in Python and evaluated using the Indian Pines, Washington DC Mall and CAVE datasets. The proposed model underwent qualitative and quantitative analysis and showed a PSNR value of 82 in the case of Indian Pines and 78.4 for the Washington DC Mall dataset at a compression rate of 10. The study proved that the proposed model provides the knowledge about complex nonlinear mapping between noise-free and noisy HSI for obtaining the denoised images and also results in high-quality compressed output.

„Miręs asmeniniame gyvenime tampu išmintingesnis pagalboje kitam“: psichoterapeutų artimųjų netekčių patyrimo reikšmė darbe su psichoterapijos klientais

The article presents the significance of the experience of bereavement for psychotherapists in their therapeutic work. This issue has not been studied much in the world and has not been raised in Lithuania yet. Paradoxically, the „psychotherapist working on his/her own“ has not been an interesting object of research so far, and so far researchers are mainly interested in the effectiveness of the psychotherapy, the question how psychotherapist feels after the crisis experience and how it affects his/her work remains in the shadows. Thus, the aim of the research presented in this paper is to uncover and systematically describe the implications of psychotherapists’ experience of bereavement for their work with psychotherapy clients. A qualitative research strategy was chosen for the study, data were collected using in-depth semistructured interviews, and analysed using Thematic Analysis. Seven psychotherapists who had experienced the loss of a loved one during the course of their psychotherapeutic work participated in the study. Three themes and seven subthemes were identified in the study, revealing the multilayered nature of experiences in their psychotherapeutic work. Although all the psychotherapists had integrated the loss of a loved one in one way or another and felt personally grown up, this had undoubtedly left traces in their therapeutic work and was experienced as a personal and professional challenge.

Struggling with fish age, a comparison of otolith preparation techniques to unravel age and growth of boarfish, Capros aper (Linnaeus, 1758)

Fish age and growth data are crucial for understanding vital species’ traits and their populations. Boarfish is a pelagic species widely distributed in the Northeast Atlantic that is one of the most discarded non-commercial species on the Portuguese coast. Due to its potential valorisation as a new fishery in the Northeast Atlantic, boarfish has become object of interest in several published studies. However, since no age standard protocol has been implemented for boarfish, studies have used a variety of different otolith preparation techniques to estimate age, leading to contradicting and discrepant results. This research aims to consolidate biological insights into boarfish age and growth, lay the foundation for a standardized ageing protocol and serve as a benchmark study for the Portuguese west coast before commercial exploitation. Through the comparison of two otolith ageing methodologies, this study recommends using whole otoliths as the best method, revealing a maximum age of 15 years. The growth pattern estimated, indicated a biphasic growth pattern, with a faster initial growth rate that slows down at 2.4 years, the moment of change, when energy is allocated for sexual maturation. This finding was corroborated by a maturity ogive and analysis of fish size and otolith length relationship.

Going Inside a Historical Brazilian Diamond from the Spada Collection (19th Century)

The characterization of objects of historical and cultural interest represents a crucial topic, specifically when it regards gemstones. Actually, the advanced investigation of precious minerals of gemological interest requires exclusively non-destructive analyses which are also suitable for determining their provenance when it is unknown. In this study, a non-destructive analytical protocol, previously tested on diamonds for petrogenetic studies, has been applied to a natural diamond of very high historical and gemological value, donated in 1852 by Monsignor Lavinio de ‘Medici Spada to the Museum of Earth Sciences of Sapienza University (Rome). The analytical protocol used includes X-ray diffraction topography, micro-computed X-ray tomography, single-crystal micro-X-ray diffraction and Fourier-transform infrared spectroscopy. The results show the presence of dislocations originating from inclusions and a very low degree of plastic deformation. The aggregation states of its N impurities show that this diamond is type IaAB, while the inclusions consist of olivine (Fo92-93), suggesting a lithospheric origin. The historical references found in the catalogs of the Museum indicate only a Brazilian origin, without any reference to the mining district. The information acquired in this study, enhanced by document research on mining in Brazil since 1700, suggests that the diamond likely comes from the district of Diamantina, Mina Gerais, Brazil.

Segmentation of image from a first-person-view unmanned aerial vehicle based on a simple ant algorithm

The object of this study is the process of image segmentation from the First Person View (FPV) of an unmanned aerial vehicle (UAV). The main hypothesis of the study assumes that the use of a simple ant algorithm could ensure the necessary quality of the segmented image. The segmentation method, unlike the known ones, takes into account the number of ants in the image, weight, initial amount and evaporation rate of the pheromone, the "greediness" of the algorithm and provides: – preliminary selection of individual channels of the Red-Green-Blue (RGB) color space; – preliminary placement of ants according to the uniform law; – determining the routes of ants; – taking into account the attractiveness of the route for each ant; – change (adjustment) in the concentration of ant pheromones; – calculation of the probability of movement (transition) of the ant on the movement route; – determination of the objective function at the j-th iteration and its minimization; – determining the coordinates of the route of movement (movement) of ants; – verification of the fulfillment of the stop condition; – determination of the best routes found by ants; – calculation of the brightness of the pixels of the segmented image in each channel of the RGB color space; – further combining the results of channel segmentation. An experimental study of image segmentation from UAV FPV based on a simple ant algorithm was conducted. The specified object of interest on the segmented image has a certain structure, unevenness of the contours, and can be further used for decoding, categorization, etc. Unlike the object of interest, the background ("garbage" objects) in the segmented image do not have a stable structure and can be further filtered out. It has been established that the segmented image by the known method based on the gradient module has a low contrast value, there are gaps in the segmented pixels of the object of interest. A segmented image using a method based on a simple ant algorithm is free from that drawback.

Protein Immobilization on Bacterial Cellulose for Biomedical Application.

New carriers for protein immobilization are objects of interest in various fields of biomedicine. Immobilization is a technique used to stabilize and provide physical support for biological micro- and macromolecules and whole cells. Special efforts have been made to develop new materials for protein immobilization that are non-toxic to both the body and the environment, inexpensive, readily available, and easy to modify. Currently, biodegradable and non-toxic polymers, including cellulose, are widely used for protein immobilization. Bacterial cellulose (BC) is a natural polymer with excellent biocompatibility, purity, high porosity, high water uptake capacity, non-immunogenicity, and ease of production and modification. BC is composed of glucose units and does not contain lignin or hemicellulose, which is an advantage allowing the avoidance of the chemical purification step before use. Recently, BC-protein composites have been developed as wound dressings, tissue engineering scaffolds, three-dimensional (3D) cell culture systems, drug delivery systems, and enzyme immobilization matrices. Proteins or peptides are often added to polymeric scaffolds to improve their biocompatibility and biological, physical-chemical, and mechanical properties. To broaden BC applications, various ex situ and in situ modifications of native BC are used to improve its properties for a specific application. In vivo studies showed that several BC-protein composites exhibited excellent biocompatibility, demonstrated prolonged treatment time, and increased the survival of animals. Today, there are several patents and commercial BC-based composites for wounds and vascular grafts. Therefore, further research on BC-protein composites has great prospects. This review focuses on the major advances in protein immobilization on BC for biomedical applications.

Evaluation of Deformable Convolution: An Investigation in Image and Video Classification

Convolutional Neural Networks (CNNs) present drawbacks for modeling geometric transformations, caused by the convolution operation’s locality. Deformable convolution (DCON) is a mechanism that solves these drawbacks and improves the robustness. In this study, we clarify the optimal way to replace the standard convolution with its deformable counterpart in a CNN model. To this end, we conducted several experiments using DCONs applied in the layers that conform a small four-layer CNN model and on the four-layers of several ResNets with depths 18, 34, 50, and 101. The models were tested in binary balanced classes with 2D and 3D data. If DCON is used on the first layers of the proposal of model, the computational resources will tend to increase and produce bigger misclassification than the standard CNN. However, if the DCON is used at the end layers, the quantity of Flops will decrease, and the classification accuracy will improve by up to 20% about the base model. Moreover, it gains robustness because it can adapt to the object of interest. Also, the best kernel size of the DCON is three. With these results, we propose a guideline and contribute to understanding the impact of DCON on the robustness of CNNs.

A Cyborg Walk for Urban Analysis? From Existing Walking Methodologies to the Integration of Machine Learning

Although walking methodologies (WMs) and machine learning (ML) have been objects of interest for urban scholars, it is difficult to find research that integrates both. We propose a ‘cyborg walk’ method and apply it to studying litter in public spaces. Walking routes are created based on an unsupervised learning algorithm (k-means) to classify public spaces. Then, a deep learning model (YOLOv5) is used to collect data from geotagged photos taken by an automatic Insta360 X3 camera worn by human walkers. Results from image recognition have an accuracy between 83.7% and 95%, which is similar to what is validated by the literature. The data collected by the machine are automatically georeferenced thanks to the metadata generated by a GPS attached to the camera. WMs could benefit from the introduction of ML for informative route optimisation and georeferenced visual data quantification. The links between these findings and the existing WM literature are discussed, reflecting on the parallels between this ‘cyborg walk’ experiment and the seminal cyborg metaphor proposed by Donna Haraway.

Geogebra in Real-Life: Important Tips to Support Student Creative Thinking in Mathematics

Creativity is an important personality trait to use in everyday life. This allows us to be flexible when facing real life situations. Mathematics education should be seen as an opportunity for creativity development, even though creativity is not traditionally associated with mathematics. One of the goals of education at every level is to encourage students to think creatively, think logically, and be able to solve problems. This study aimed to provide important tips to support the student creative thinking when studying mathematics. This study also described activities outside the classroom that could develop the student creativity related to geometry. The methods used were surveys, interviews, and classroom and outdoor observations. We used brainstorming in mathematics education. The results of the research during the survey were visible. Researchers collaborated with students aged 19 to 21 years. We asked them to search for “any geometry” and take pictures of interesting objects while they were doing the activity. Next, students created ideas for geometry assignments based on the photos they took. The large collection of photos collected represented the basis for creating mathematical problems without any specific steps related to geometric objects. These open problems could be solved by students. In conclusion, Geogebra could encourage the student creative thinking by providing the opportunity to experiment with their own ideas and find solutions to mathematical problems.

Direction-aware multi-branch attention and Gaussian label assignment for remote sensing aggregative object detection

ABSTRACT A large number of objects of different scales appear aggregative distribution in remote sensing images, which brings two challenges to remote sensing object detection. The first challenge arises from the presence of objects with large scale differences within the same scene, which complicates the detector’s ability to balance the detection performance between large and small objects. The second challenge is that the distribution of dense objects on remote sensing images with complex backgrounds can reduce the feature representation capability of the network, leading to false detections and missed detections of objects. To address the two challenges present in remote sensing object detection, we propose a new object detection network DA-GLANet (Direction-Aware Multi-Branch Attention and Gaussian Label Assignment Balance Net). For the first challenge, we designed a multi-scale feature fusion module to fully utilize semantic information at different scales, thereby improving the detection accuracy of multi-scale objects. Subsequently, to tackle the second challenge, we proposed a multi-branch attention module with direction-aware capabilities. This module can enhance the network’s feature representation of objects of interest by generating attention maps that are both direction-aware and position-sensitive. In addition, we propose a label assignment strategy based on Gaussian distribution to address the imbalance of positive and negative labels caused by the complexity of remote sensing images. These methods work synergistically to improve aggregative object detection accuracy. We have conducted extensive experiments on the DOTA dataset, Levir dataset, and WHU buildings change detection dataset to demonstrate the superiority of our method. On the DOTA dataset, our method improves mAP by 0.4%, 0.3%, and 0.2% compared to the top three methods. On the Levir dataset, our method improves mAP by 2.5%, 2.2%, and 1.6% compared to the top three methods.