Explore the vast range of titles available.

Anomaly detection in video surveillance is a highly developed subject that is attracting increased attention from the research community. There is great demand for intelligent systems with the capacity to automatically detect anomalous events in streaming videos. Due to this, a wide variety of approaches have been proposed to build an effective model that would ensure public security. There has been a variety of surveys of anomaly detection, such as of network anomaly detection, financial fraud detection, human behavioral analysis, and many more. Deep learning has been successfully applied to many aspects of computer vision. In particular, the strong growth of generative models means that these are the main techniques used in the proposed methods. This paper aims to provide a comprehensive review of the deep learning-based techniques used in the field of video anomaly detection. Specifically, deep learning-based approaches have been categorized into different methods by their objectives and learning metrics. Additionally, preprocessing and feature engineering techniques are discussed thoroughly for the vision-based domain. This paper also describes the benchmark databases used in training and detecting abnormal human behavior. Finally, the common challenges in video surveillance are discussed, to offer some possible solutions and directions for future research.

CRISPR-associated transposons (CAST) are programmable mobile genetic elements that insert large DNA cargos using an RNA-guided mechanism 1 – 3 . CAST elements contain multiple conserved proteins: a CRISPR effector (Cas12k or Cascade), a AAA+ regulator (TnsC), a transposase (TnsA–TnsB) and a target-site-associated factor (TniQ). These components are thought to cooperatively integrate DNA via formation of a multisubunit transposition integration complex (transpososome). Here we reconstituted the approximately 1 MDa type V-K CAST transpososome from Scytonema hofmannii ( Sh CAST) and determined its structure using single-particle cryo-electon microscopy. The architecture of this transpososome reveals modular association between the components. Cas12k forms a complex with ribosomal subunit S15 and TniQ, stabilizing formation of a full R-loop. TnsC has dedicated interaction interfaces with TniQ and TnsB. Of note, we observe TnsC–TnsB interactions at the C-terminal face of TnsC, which contribute to the stimulation of ATPase activity. Although the TnsC oligomeric assembly deviates slightly from the helical configuration found in isolation, the TnsC-bound target DNA conformation differs markedly in the transpososome. As a consequence, TnsC makes new protein–DNA interactions throughout the transpososome that are important for transposition activity. Finally, we identify two distinct transpososome populations that differ in their DNA contacts near TniQ. This suggests that associations with the CRISPR effector can be flexible. This Sh CAST transpososome structure enhances our understanding of CAST transposition systems and suggests ways to improve CAST transposition for precision genome-editing applications. Structural studies of the CRISPR-associated transposon comprising Cas12k, TnsC, TnsB and TniQ from Scytonema hofmannii using cryo-electron microscopy reveal insights into the architecture and mechanism of RNA-guided DNA transposition.

Photochemical reactions are a powerful tool in (bio)materials design due to the spatial and temporal control light can provide. To extend their applications in biological setting, the use of low-energy, long wavelength light with high penetration propertiesis required. Further regulation of the photochemical process by additional stimuli, such as pH, will open the door for construction of highly regulated systems in nanotechnology- and biology-driven applications. Here we report the green light induced [2+2] cycloaddition of a halochromic system based on a styrylquinoxaline moiety, which allows for its photo-reactivity to be switched on and off by adjusting the pH of the system. Critically, the [2+2] photocycloaddition can be activated by green light (λ up to 550 nm), which is the longest wavelength employed to date in catalyst-free photocycloadditions in solution. Importantly, the pH-dependence of the photo-reactivity was mapped by constant photon action plots. The action plots further indicate that the choice of solvent strongly impacts the system’s photo-reactivity. Indeed, higher conversion and longer activation wavelengths were observed in water compared to acetonitrile under identical reaction conditions. The wider applicability of the system was demonstrated in the crosslinking of an 8-arm PEG to form hydrogels (ca. 1 cm in thickness) with a range of mechanical properties and pH responsiveness, highlighting the potential of the system in materials science. Light gated reactions are important due to their spatial and temporal control over the chemical processes but long wavelength activation of photocycloaddition reactions are rare. Here the authors introduce a green light induced [2+2] cycloaddition of a halochromic system, which allows for its photo-reactivity to be switched on and off by adjusting the pH of the system.

Following the introduction of data privacy regulations and “the right to be forgotten”, large language models (LLMs) unlearning has emerged as a promising data removal solution for compliance purposes, while also facilitating a diverse range of applications, including copyright protection, model detoxification and correction, and jailbreaking defence. In this survey, we present the taxonomy of existing LLMs unlearning algorithms, summarise unlearning evaluation methods including specialised benchmarks and threat models, and explore the applications of unlearning to provide a broad overview of the current state-of-the-art. We propose a novel problem formulation of LLMs unlearning with the additional unlearning objective: “robustness” to reflect the growing research interest in not only effectively and efficiently eliminating unwanted data, but also ensuring the process is performed safely and securely. To the best of our knowledge, we are the first to examine the robustness of unlearning algorithms as well as threat models for robustness evaluation, aspects that have not been assessed in past surveys. We also identify the limitations of the current approaches, including limited applicability to black-box models, vulnerability to adversarial attacks and knowledge leakage, and inefficiency, all of which require further improvement in future works. Furthermore, our survey highlights future directions for LLMs unlearning research, such as the development of comprehensive evaluation benchmarks, the movement towards robust unlearning and explainable AI for unlearning mechanisms, and addressing potential ethical dilemmas in unlearning governance.

Estimating tropical cyclone (TC) intensity is crucial for disaster reduction and risk management. This study aims to estimate TC intensity using machine learning (ML) models. We utilized eight ML models to predict TC intensity, incorporating factors such as TC location, central pressure, distance to land, landfall in the next six hours, storm speed, storm direction, date, and number from the International Best Track Archive for Climate Stewardship Version 4 (IBTrACS V4). The dataset was divided into four sub-datasets based on the El Niño–Southern Oscillation (ENSO) phases (Neutral, El Niño, and La Niña). Our results highlight that central pressure has the greatest effect on TC intensity estimation, with a maximum root mean square error (RMSE) of 1.289 knots (equivalent to 0.663 m/s). Cubist and Random Forest (RF) models consistently outperformed others, with Cubist showing superior performance in both training and testing datasets. The highest bias was observed in SVM models. Temporal analysis revealed the highest mean error in January and November, and the lowest in February. Errors during the Warm phase of ENSO were notably higher, especially in the South China Sea. Central pressure was identified as the most influential factor for TC intensity estimation, with further exploration of environmental features recommended for model robustness.

There have been few independent evaluations of computer-aided detection (CAD) software for tuberculosis (TB) screening, despite the rapidly expanding array of available CAD solutions. We developed a test library of chest X-ray (CXR) images which was blindly re-read by two TB clinicians with different levels of experience and then processed by 12 CAD software solutions. Using Xpert MTB/RIF results as the reference standard, we compared the performance characteristics of each CAD software against both an Expert and Intermediate Reader, using cut-off thresholds which were selected to match the sensitivity of each human reader. Six CAD systems performed on par with the Expert Reader (Qure.ai, DeepTek, Delft Imaging, JF Healthcare, OXIPIT, and Lunit) and one additional software (Infervision) performed on par with the Intermediate Reader only. Qure.ai, Delft Imaging and Lunit were the only software to perform significantly better than the Intermediate Reader. The majority of these CAD software showed significantly lower performance among participants with a past history of TB. The radiography equipment used to capture the CXR image was also shown to affect performance for some CAD software. TB program implementers now have a wide selection of quality CAD software solutions to utilize in their CXR screening initiatives.

Yield estimation (YE) of the crop is one of the main tasks in fruit management and marketing. Based on the results of YE, the farmers can make a better decision on the harvesting period, prevention strategies for crop disease, subsequent follow-up for cultivation practice, etc. In the current scenario, crop YE is performed manually, which has many limitations such as the requirement of experts for the bigger fields, subjective decisions and a more time-consuming process. To overcome these issues, an intelligent YE system was proposed which detects, localizes and counts the number of tomatoes in the field using SegNet with VGG19 (a deep learning-based semantic segmentation architecture). The dataset of 672 images was given as an input to the SegNet with VGG19 architecture for training. It extracts features corresponding to the tomato in each layer and detection was performed based on the feature score. The results were compared against the other semantic segmentation architectures such as U-Net and SegNet with VGG16. The proposed method performed better and unveiled reasonable results. For testing the trained model, a case study was conducted in the real tomato field at Manapparai village, Trichy, India. The proposed method portrayed the test precision, recall and F1-score values of 89.7%, 72.55% and 80.22%, respectively along with reasonable localization capability for tomatoes.

This study investigates landslide occurrence in Taiwan, a region highly susceptible to landslides due to steep mountains and frequent typhoons (TYPs). The primary objective is to understand how both geomorphological factors and TYP characteristics contribute to landslide occurrence, which is essential for improving hazard prediction and risk management. The research analyzed landslide events that occurred during the TYP seasons of 2019 and 2023. The methodology involved using satellite-derived landslide inventories from SPOT imagery for events larger than 0.1 hectares, tropical cyclone track and intensity data from IBTrACS v4 (classified by Saffir–Simpson Hurricane Scale), and detailed topographic variables (elevation, slope, aspect, Stream Power Index) extracted from a 30 m Shuttle Radar Topography Mission Digital Elevation Model (SRTM-DEM). Land use and land cover classifications were based on Landsat imagery. To establish a timeline, landslides were matched with TYPs within a ±3-day window, and proximity was analyzed using buffer zones ranging from 50 to 500 km around storm centers. Key findings revealed that landslide susceptibility results from a complex interplay of meteorological, topographic, and land cover factors. The critical controls identified include elevations above 2000 m, slope angles between 30 and 45 degrees, southeast- and south-facing aspects, and low Stream Power Index values typical of headwater and upper slope locations. Landslides were most frequent during Category 3 TYPs and were concentrated 300 to 350 km from storm centers, where optimal rainfall conditions for slope failures exist. Interestingly, despite the stronger storms in 2023, the number of landslides was higher in 2019. This emphasizes the importance of interannual variability and terrain preparedness. These findings support sustainable disaster risk reduction and climate-resilient development, aligning with Sustainable Development Goals 11 (Sustainable Cities and Communities) and 13 (Climate Action). Furthermore, they provide a foundation for improving hazard assessment and risk mitigation in Taiwan and similar mountainous, TYP-prone regions.

Thioindigos are visible light responsive photoswitches with excellent spatial control over the conformational change between their trans- and cis- isomers. However, they possess limited solubility in all conventional organic solvents and polymers, hindering their application in soft matter materials. Herein, we introduce a strategy for the covalent insertion of thioindigo units into polymer main chains, enabling thioindigos to function within crosslinked polymeric hydrogels. We overcome their solubility issue by developing a thioindigo bismethacrylate linker able to undergo radical initiated thiol-ene reaction for step-growth polymerization, generating indigo-containing polymers. The optimal wavelength for the reversible trans- / cis- isomerisation of thioindigo was elucidated by constructing a detailed photochemical action plot of their switching efficiencies at a wide range of monochromatic wavelengths. Critically, indigo-containing polymers display significant photoswitching of the materials’ optical and physical properties in organic solvents and water. Furthermore, the photoswitching of thioindigo within crosslinked structures enables visible light induced modulation of the hydrogel stiffness. Both the thioindigo-containing hydrogels and photoswitching processes are non-toxic to cells, thus offering opportunities for advanced applications in soft matter materials and biology-related research. Thioindigos are reversible photoswitches with spatial control over the conformational change, yet have very limited solubility in most solvents. Here, the authors report a method for the insertion of thioindigos into polymer chains, allowing the formation of visible light responsive hydrogels.

Metformin treatment for 10 d rescues several behavioral deficits related to fragile X syndrome in adult Fmr1 −/y mice. Fragile X syndrome (FXS) is the leading monogenic cause of autism spectrum disorders (ASD). Trinucleotide repeat expansions in FMR1 abolish FMRP expression, leading to hyperactivation of ERK and mTOR signaling upstream of mRNA translation. Here we show that metformin, the most widely used drug for type 2 diabetes, rescues core phenotypes in Fmr1 −/y mice and selectively normalizes ERK signaling, eIF4E phosphorylation and the expression of MMP-9. Thus, metformin is a potential FXS therapeutic.