Generation Method Research Articles

Structure-based drug design with equivariant diffusion models.

Structure-based drug design (SBDD) aims to design small-molecule ligands that bind with high affinity and specificity to pre-determined protein targets. Generative SBDD methods leverage structural data of drugs with their protein targets to propose new drug candidates. However, most existing methods focus exclusively on bottom-up de novo design of compounds or tackle other drug development challenges with task-specific models. The latter requires curation of suitable datasets, careful engineering of the models and retraining from scratch for each task. Here we show how a single pretrained diffusion model can be applied to a broader range of problems, such as off-the-shelf property optimization, explicit negative design and partial molecular design with inpainting. We formulate SBDD as a three-dimensional conditional generation problem and present DiffSBDD, an SE(3)-equivariant diffusion model that generates novel ligands conditioned on protein pockets. Furthermore, we show how additional constraints can be used to improve the generated drug candidates according to a variety of computational metrics.

Systematic Review of Generative Modelling Tools and Utility Metrics for Fully Synthetic Tabular Data

Sharing data with third parties is essential for advancing science, but it is becoming more and more difficult with the rise of data protection regulations, ethical restrictions, and growing fear of misuse. Fully synthetic data, which transcends anonymisation, may be the key to unlocking valuable untapped insights stored away in secured data vaults. This review examines current synthetic data generation methods and their utility measurement. We found that more traditional generative models such as Classification and Regression Tree models alongside Bayesian Networks remain highly relevant and are still capable of surpassing deep learning alternatives like Generative Adversarial Networks. However, our findings also display the same lack of agreement on metrics for evaluation, uncovered in earlier reviews, posing a persistent obstacle to advancing the field. We propose a tool for evaluating the utility of synthetic data and illustrate how it can be applied to three synthetic data generation models. By streamlining evaluation and promoting agreement on metrics, researchers can explore novel methods and generate compelling results that will convince data curators and lawmakers to embrace synthetic data. Our review emphasises the potential of synthetic data and highlights the need for greater collaboration and standardisation to unlock its full potential.

Generative Design Method for Single-Layer Spatial Grid Structural Joints

Single-layer spatial grid joints are crucial to structural safety, with commonly used welded hollow spherical joints and cast steel joints. However, these traditional joints face limitations, including a rigid design, excessive weight, and susceptibility to stress concentration. As engineering practices advance, these joints struggle to meet modern requirements. This paper introduces a generative method for designing rigid joints in single-layer spatial grid structures, based on Audze space-filling criteria. The method’s mathematical formulation is presented, followed by developing novel joint configurations by exploring various cross-sectional forms, retention mass, and geometric elements, while considering bending moments. A comparative analysis of static properties between the new and traditional joints shows promising results. The generative approach demonstrates significant innovation, producing lightweight, aesthetically pleasing, and structurally efficient joints. Compared to conventional welded hollow spherical joints, the new joints exhibit a 57% reduction in self-weight, a 51% decrease in maximum equivalent stress, and a 24% reduction in maximum displacement. This method enables versatile and optimized joint design for single-layer spatial grid structures, offering enhanced strength, safety, and aesthetic appeal.

Using UMAP for Partially Synthetic Healthcare Tabular Data Generation and Validation

In healthcare, vast amounts of data are increasingly collected through sensors for smart health applications and patient monitoring or diagnosis. However, such medical data often comprise sensitive patient information, posing challenges regarding data privacy, and are resource-intensive to acquire for significant research purposes. In addition, the common case of lack of information due to technical issues, transcript errors, or differences between descriptors considered in different health centers leads to the need for data imputation and partial data generation techniques. This study introduces a novel methodology for partially synthetic tabular data generation, designed to reduce the reliance on sensor measurements and ensure secure data exchange. Using the UMAP (Uniform Manifold Approximation and Projection) visualization algorithm to transform the original, high-dimensional reference data set into a reduced-dimensional space, we generate and validate synthetic values for incomplete data sets. This approach mitigates the need for extensive sensor readings while addressing data privacy concerns by generating realistic synthetic samples. The proposed method is validated on prostate and breast cancer data sets, showing its effectiveness in completing and augmenting incomplete data sets using fully available references. Furthermore, our results demonstrate superior performance in comparison to state-of-the-art imputation techniques. This work makes a dual contribution by not only proposing an innovative method for synthetic data generation, but also studying and establishing a formal framework to understand and solve synthetic data generation and imputation problems in sensor-driven environments.

Improving Circulating Tumor Cell Detection Using Image Synthesis and Transformer Models in Cancer Diagnostics

Cancer is the second leading cause of death, significantly threatening human health. Effective treatment options are often lacking in advanced stages, making early diagnosis crucial for reducing mortality rates. Circulating tumor cells (CTCs) are a promising biomarker for early detection; however, their automatic detection is challenging due to their heterogeneous size and shape, as well as their scarcity in blood. This study proposes a data generation method using the Segment Anything Model (SAM) combined with a copy–paste strategy. We develop a detection network based on the Swin Transformer, featuring a backbone network, scale adapter module, shape adapter module, and detection head, which enhances CTC localization and identification in images. To effectively utilize both generated and real data, we introduce an improved loss function that includes a regularization term to ensure consistency across different data distributions. Our model demonstrates exceptional performance across five evaluation metrics: accuracy (0.9960), recall (0.9961), precision (0.9804), specificity (0.9975), and mean average precision (mAP) of 0.9400 at an Intersection over Union (IoU) threshold of 0.5. These results are achieved on a dataset generated by mixing both public and local data, highlighting the robustness and generalizability of the proposed approach. This framework surpasses state-of-the-art models (ADCTC, DiffusionDet, CO-DETR, and DDQ), providing a vital tool for early cancer diagnosis, treatment planning, and prognostic assessment, ultimately enhancing human health and well-being.

Tactical intent-driven autonomous air combat behavior generation method

With the rapid development and deep application of artificial intelligence, modern air combat is incrementally evolving towards intelligent combat. Although deep reinforcement learning algorithms have contributed to dramatic advances in in air combat, they still face challenges such as poor interpretability and weak transferability of adversarial strategies. In this regard, this paper proposes a tactical intent-driven method for autonomous air combat behaviour generation. Firstly, this paper explores the mapping relationship between optimal strategies and rewards, demonstrating the detrimental effects of the combination of sparse rewards and dense rewards on policy. Built around this, the decision-making process of pilot behavior is analyzed, and a reward mapping model from intent to behavior is established. Finally, to address the problems of poor stability and slow convergence speed of deep reinforcement learning algorithms in large-scale state-action spaces, the dueling-noisy-multi-step DQN algorithm is devised, which not only improves the accuracy of value function approximation but also enhances the efficiency of space exploration and network generalization. Through experiments, the conflicts between sparse rewards and dense rewards are demonstrated. The superior performance and stability of the proposed algorithm compared to other algorithms are captured by our empirical results. More intuitively, the strategies under different intents exhibit strong interpretability and flexibility, which can provide tactical support for intelligent decision-making in air combat.

CSearch: chemical space search via virtual synthesis and global optimization

The two key components of computational molecular design are virtually generating molecules and predicting the properties of these generated molecules. This study focuses on an effective method for molecular generation through virtual synthesis and global optimization of a given objective function. Using a pre-trained graph neural network (GNN) objective function to approximate the docking energies of compounds for four target receptors, we generated highly optimized compounds with 300–400 times less computational effort compared to virtual compound library screening. These optimized compounds exhibit similar synthesizability and diversity to known binders with high potency and are notably novel compared to library chemicals or known ligands. This method, called CSearch, can be effectively utilized to generate chemicals optimized for a given objective function. With the GNN function approximating docking energies, CSearch generated molecules with predicted binding poses to the target receptors similar to known inhibitors, demonstrating its effectiveness in producing drug-like binders.Scientific Contribution We have developed a method for effectively exploring the chemical space of drug-like molecules using a global optimization algorithm with fragment-based virtual synthesis. The compounds generated using this method optimize the given objective function efficiently and are synthesizable like commercial library compounds. Furthermore, they are diverse, novel drug-like molecules with properties similar to known inhibitors for target receptors.

Rapid Vehicle Trajectory Prediction Based on Multi-Attention Mechanism for Fusing Multimodal Information

Trajectory prediction plays a crucial role in autonomous driving tasks, as accurately and rapidly predicting the future trajectories of traffic participants can significantly enhance the safety and robustness of autonomous driving systems. This paper presents a novel trajectory prediction model that follows the encoder–decoder paradigm, achieving precise and rapid predictions of future vehicle trajectories by efficiently aggregating the spatiotemporal and interaction information of agents in traffic scenarios. We propose an agent–agent interaction information extraction module based on a sparse graph attention mechanism, which enables the efficient aggregation of interaction information between agents. Additionally, we introduce a non-autoregressive query generation method that accelerates the model inference speed by generating the decoding queries in parallel. Comparative experiments with the existing advanced algorithms show that our method improves the multimodal trajectory prediction metrics for the Minimum Average Displacement Error (minADE), the Minimum Final Displacement Error (minFDE), and the Miss Rate (MR) by an average of 9.1%, 11.8%, and 14.6%, respectively, while the inference time is only 33.7% of the average time taken by the other algorithms. Finally, we demonstrate the effectiveness of the various modules proposed in this paper through ablation studies.

A new generative method for multi-focus image fusion of underwater micro bubbles

The optical detection methodology stands as a predominant approach for detecting underwater bubbles. Nonetheless, owing to poor underwater imaging conditions, the acquired image depth of field proves inadequate, posing significant challenges for the study and identification of underwater micro bubbles. In this investigation, we present a multi-focus image fusion model tailored for underwater micro bubbles, grounded in the Denoising Diffusion Probabilistic Model. We also propose a multi-focus image fusion metric suitable for underwater scenarios with micro bubbles. Experimental validation on the constructed dataset demonstrates that our model achieves better results than traditional methods. These results substantiate the model’s efficacy in conserving image characteristics and attaining multi-focus fusion. Consequently, this research furnishes substantial empirical support for subsequent endeavors in image-related tasks.

Channel Attention-Based Conditional Diffusion Model Applied to Fault Diagnosis Under Imbalanced Data

Issues such as data scarcity and data imbalance have long posed significant difficulties in the field of intelligent fault diagnosis. They lead to reduced diagnostic accuracy and endanger the safety and reliability of industrial equipment. To address these challenges, this study introduces a novel channel attention-based conditional diffusion model (CAC-DM) that recalibrates features through a squeeze-and-excitation process. This enhancement boosts the model’s ability to focus on critical features while suppressing irrelevant information, thereby improving the UNet network’s discrimination capability in handling small-sample faults that are highly similar in nature. Experimental validation demonstrates that CAC-DM performs exceptionally well in scenarios with high class similarity, effectively distinguishing among categories with similar distributions in limited data and generating high-quality samples. Compared to existing generative methods, the CAC-DM exhibits significant advantages in producing distinguishable fault samples, particularly in cases of sample imbalance. This approach offers an effective new solution for fault diagnosis.

Harnessing Ruthenium and Copper Catalysts for Formate Dehydrogenase Reactions.

Formic acid (HCOOH) is a promising source of hydrogen energy that can be used to produce hydrogen in a more economical and ecological way. Formic acid is a simple carboxylic acid with a high hydrogen concentration and is generally stable, making it useful as a hydrogen transporter. Catalytic dehydrogenation is usually used to extract hydrogen from formic acid; this process releases hydrogen gas and yields carbon dioxide as a byproduct. Comparing this technology to conventional hydrogen generation methods, there are several benefits, such as the utilization of the formic acid handling infrastructure already in place and the possibility of a simpler integration into different energy systems. Notwithstanding, several obstacles persist, including enhancing the effectiveness of the dehydrogenation procedure and reducing the ecological consequences of the correlated carbon dioxide discharges. Catalysts, reaction conditions, and carbon collection and utilization methodologies are all being researched further. The development of Ru and Cu-based catalysts for the catalytic breakdown of HCOOH into CO2 and H2 is the main topic of this account. Herein, the focus is on the kinetic studies of HCOOH dehydrogenation, encompassing mechanistic investigations that consider intermediate studies and DFT calculations.

A Study on Wind Collection Effect of Vertical Axis Windmills

In recent years, global warming caused by greenhouse gasses such as carbon dioxide has become a concern. This has resulted in increased focus on environmentally friendly power systems. Consequently, renewable energy power generation methods, such as wind and solar power generation, have attracted attention. Wind power generation is expected to significantly increase in the future. However, in many inland areas in Japan, the average wind speed remains 6 m/s or less. In this study, we proposed the introduction of winglets and wind collectors (used in aircraft wings) into straight-wing vertical-axis wind turbines to improve their power generation efficiency. Field tests were conducted to confirm the effectiveness of the proposed method. Using winglets and wind collectors, the wind turbine rotation speed was increased at low wind speeds, which facilitated the generation of power. Moreover, it was confirmed that a wind turbine equipped with the proposed winglets and wind collectors could capture wind without its dispersal as it passed through the turbine.

Generative Artificial Intelligence in the Early Diagnosis of Gastrointestinal Disease

This study reviews the recent progress of generative artificial intelligence for gastrointestinal disease (GID) from detection to diagnosis. The source of data was 16 original studies in PubMed. The search terms were ((gastro* [title]) or (endo* [title])) and ((GAN [title/abstract] or (transformer [title/abstract]). The eligibility criteria were as follows: (1) the dependent variable of gastrointestinal disease; (2) the interventions of generative adversarial network (GAN) and/or transformer for classification, detection and/or segmentation; (3) the outcomes of accuracy, intersection of union (IOU), structural similarity and/or Dice; (3) the publication period of 2021–2023; (4) the publication language of English. Based on the results of this study, different generative artificial intelligence methods would be appropriate for different tasks for the early diagnosis of gastrointestinal disease. For example, patch GAN (accuracy 91.9%) in the case of classification, bi-directional cycle GAN (structural similarity 98.8%) in the case of data generation and semi-supervised GAN (Dice 89.4%) in the case of segmentation. Their performance indicators reported varied within 87.1–91.9% for accuracy, 83.0–98.8% for structural similarity and 86.6–89.4% for Dice. Likewise, vision transformer (accuracy 96.9%) in the case of classification, multi-modal transformer (IOU 79.5%) in the case of detection and multi-modal transformer (Dice 89.5%) in the case of segmentation. Their performance measures reported registered a variation within 85.7–96.9% for accuracy, 79.5% for IOU and 77.8–89.5% for Dice. Synthesizing different kinds of generative artificial intelligence for different kinds of GID data would further the horizon of research on this topic. In conclusion, however, generative artificial intelligence provides an effective, non-invasive decision support system for the early diagnosis of gastrointestinal disease from detection to diagnosis.

On Rotor Profling of Internally Geared Screw Machines

Abstract The internally geared screw machine represents a novel type of positive displacement compressor which consists of an inner and outer rotor. Both rotors rotate in the same direction but are each centered on offset parallel axes. The rotor profiles are designed to create multiple continuous contact points between the rotors, forming several separate working chambers whose volumes vary from minimum to maximum and back to minimum during a single rotation of the outer rotor. For a gas or two-phase working fluid, adjusting the discharge port geometry allows internal compression to occur before discharge. Previous research has focused on using the well-known rotor profling method, which employs a circular pin to generate the inner and outer rotor profiles. Although the rack method for rotor profile generation has been described and investigated for conventional screw machine rotor profiles, it has never been applied to internally geared screw machine profile generation. This paper provides an initial description and application of the rack method for generating internally geared screw machine profiles. Potential benefits of using the rack method compared to conventional methods for rotor profile generation in internally geared machines are discussed. Additionally, the limitations of using the rack method for internal gearing are presented and illustrated through various examples and applications.

Text2NeRF: Text-Driven 3D Scene Generation With Neural Radiance Fields.

Text-driven 3D scene generation is widely applicable to video gaming, film industry, and metaverse applications that have a large demand for 3D scenes. However, existing text-to-3D generation methods are limited to producing 3D objects with simple geometries and dreamlike styles that lack realism. In this work, we present Text2NeRF, which is able to generate a wide range of 3D scenes with complicated geometric structures and high-fidelity textures purely from a text prompt. To this end, we adopt NeRF as the 3D representation and leverage a pre-trained text-to-image diffusion model to constrain the 3D reconstruction of the NeRF to reflect the scene description. Specifically, we employ the diffusion model to infer the text-related image as the content prior and use a monocular depth estimation method to offer the geometric prior. Both content and geometric priors are utilized to update the NeRF model. To guarantee textured and geometric consistency between different views, we introduce a progressive scene inpainting and updating strategy for novel view synthesis of the scene. Our method requires no additional training data but only a natural language description of the scene as the input. Extensive experiments demonstrate that our Text2NeRF outperforms existing methods in producing photo-realistic, multi-view consistent, and diverse 3D scenes from a variety of natural language prompts. Our code and model are available at https://github.com/eckertzhang/Text2NeRF.

Random Permutation Set Reasoning.

In artificial intelligence, it is crucial for pattern recognition systems to process data with uncertain information, necessitating uncertainty reasoning approaches such as evidence theory. As an orderable extension of evidence theory, random permutation set (RPS) theory has received increasing attention. However, RPS theory lacks a suitable generation method for the element order of permutation mass function (PMF) and an efficient determination method for the fusion order of permutation orthogonal sum (POS). To solve these two issues, this paper proposes a reasoning model for RPS theory, called random permutation set reasoning (RPSR). RPSR consists of three techniques, including RPS generation method (RPSGM), RPSR rule of combination, and ordered probability transformation (OPT). Specifically, RPSGM can construct RPS based on Gaussian discriminant model and weight analysis; RPSR rule incorporates POS with reliability vector, which can combine RPS sources with reliability in fusion order; OPT is used to convert RPS into a probability distribution for the final decision. Besides, numerical examples are provided to illustrate the proposed RPSR. Moreover, the proposed RPSR is applied to classification problems. An RPSR-based classification algorithm (RPSRCA) and its hyperparameter tuning method are presented. The results demonstrate the efficiency and stability of RPSRCA compared to existing classifiers.

Deepfakes in digital media forensics: Generation, AI-based detection and challenges

Deepfake technology presents significant challenges for digital media forensics. As deepfakes become increasingly sophisticated, the ability to detect and attribute manipulated media becomes more difficult. The main challenge lies in the realistic and convincing nature of deepfakes, which can deceive human perception and traditional forensic techniques. Furthermore, the widespread availability of open-source deepfake tools and increasing computational power contribute to the ease with which malicious actors can create and disseminate deepfakes. The challenges posed by deepfakes for digital media forensics are multifaceted. Therefore, the development of sophisticated detection algorithms, the creation of comprehensive datasets, and the establishment of legal frameworks are crucial in addressing these challenges. This paper provides a comprehensive analysis of current methods for deepfake generation and the issues surrounding their detection. It also explores the potential of modern AI-based detection techniques in combating the proliferation of deepfakes. This analysis aims to contribute to advancing deepfake detection by highlighting the limits of current detection techniques, the most relevant issues, the upcoming challenges, and suggesting future directions for research.

Semantic Mask Reconstruction and Category Semantic Learning for few-shot image generation

Few-shot image generation aims at generating novel images for the unseen category when given K images from the same category. Despite significant advancements in existing few-shot image generation methods, great challenges remain regarding the quality and diversity of the generated images. This issue stems from the model’s struggle to fully comprehend the semantic content of images and extract sufficiently semantic representations. To address these issues, we propose a semantic mask reconstruction (SMR) and category semantic learning (CSL) method for few-shot image generation. Specifically, SMR performs mask reconstruction in a high-level semantic space and designs a strategy for dynamically adjusting the mask ratio, which increases the difficulty of the generation tasks by gradually increasing the mask ratio to enhance the learning ability of the discriminator, thereby prompting the generator to learn more critical features relevant to the generation task. In addition, CSL introduces a triplet loss to optimize the distance between the generated image, its corresponding input image, and input images of other categories. This encourages the generative model to discern subtle differences between categories, thereby achieving more fine-grained generation and improving the fidelity of generated images. Both SMR and CSL can function as plug-and-play modules. Extensive experimental results across three standard datasets demonstrate that the SMR-CSL outperforms other methods in terms of the quality and diversity of the generated images. Furthermore, the results of downstream classification experiments verify that the images generated by the proposed method can effectively assist downstream classification tasks.

The seventh blind test highlights exciting developments in crystal structure prediction.

Two reports on the seventh blind test on crystal structure prediction extensively discuss the cutting-edge avant-garde methods of structure generation and energy ranking.

Parametric analysis of self-excited aeroelastic instability of an isolated single-axis two-dimensional flat solar tracker

The escalating global energy demand is driving a transition towards sustainable energy sources, particularly solar power, which is experiencing significant growth. Single-axis, horizontally mounted photovoltaic (PV) solar trackers are emerging as one of the most cost-effective methods for solar energy generation. However, the reduction in structural stiffness has led to an increase in aeroelastic issues, some of which have resulted in catastrophic failures around the world, posing significant challenges. Recent aerodynamic studies have attributed these instabilities to stall flutter. The absence of a validated theory hampers early estimation of these aeroelastic phenomena. The primary aim in this work is to develop a semi-empirical framework to predict aeroelastic instabilities in isolated two-dimensional solar trackers, focusing on dynamic self-excited responses. This effort aims to improve the reliability and efficiency of solar tracker designs, optimizing energy utilization in the context of the evolving renewable energy landscape. The semi-empirical formulation developed here reveals the influence of geometric and mechanical characteristics on the critical wind speed at which aeroelastic instabilities set in. The effects of reduced inertia (m⁎) and mechanical damping coefficient (ξmech) of the structure on the critical wind speed at which aeroelastic instabilities arise are analyzed here. It is shown that increasing either inertia or damping has a minor impact on the critical speed when the initial values of these parameters are large, while the changes are significant when these initial values are small.