Generating high-quality, controllable, and structurally consistent 3D scenes in complex multi-object environments remains a fundamental challenge. We present SceneGenesis, a unified framework that synthesizes 3D scenes by combining semantic structural priors with mesh-guided video–geometry fusion. SceneGenesis first employs large language models to convert textual descriptions into category-aware object specifications, which are transformed into structured meshes using procedural approximations and pretrained asset generators, enabling precise layout control and scalable scene construction. To obtain rich and style-controllable appearances, SceneGenesis generates multi-view video representations conditioned on the initialized structure. A mesh-guided video–geometry fusion module then consolidates video evidence with mesh priors through mesh-conditioned fragment initialization, progressive geometric refinement, and structure-aware optimization, substantially improving global geometric fidelity and visual realism. Experiments demonstrate that SceneGenesis supports flexible style variation and object-level editing while achieving strong controllability, scalability, and structural quality.

Recent advances in text-to-3D creation integrate the potent prior of Diffusion Models from text-to-image generation into 3D domain. Nevertheless, generating 3D scenes with multiple objects remains challenging. Therefore, we present DreamScape, a method for generating 3D scenes from text. Utilizing Gaussian Splatting for 3D representation, DreamScape introduces 3D Gaussian Guide that encodes semantic primitives, spatial transformations and relationships from text using LLMs, enabling local-to-global optimization. Progressive scale control is tailored during local object generation, addressing training instability issue arising from simple blending in the global optimization stage. Collision relationships between objects are modeled at the global level to mitigate biases in LLMs priors, ensuring physical correctness. Additionally, to generate pervasive objects like rain and snow distributed extensively across the scene, we design specialized sparse initialization and densification strategy. Experiments demonstrate that DreamScape achieves state-of-the-art performance, enabling high-fidelity, controllable 3D scene generation.

A robust estimation method, Balanced Least Absolute Value Estimator (BLAVE), is introduced and compared with the traditional RANdom SAmple Consensus (RANSAC) method. The comparison is performed empirically by applying both estimators on the camera motion parameters estimation problem. A linearised model for this estimation problem is derived. The tests were performed on a simulated scene with added random noise and gross errors as well as on actual images taken by a mobile mapping system. The greatest advantage of BLAVE is that it processes all observations at once as well as its median‐like property: the estimated parameters are not influenced by the size of the outliers. It can tolerate up to 50% outliers in data and still produce accurate results. The greatest disadvantage of RANSAC is that the results are not repeatable because of the random sampling of data. Moreover, the results are less accurate, because RANSAC generally does not produce a ‘best‐fit’ parameter estimation. The number of trials, which must be tested by RANSAC to find a reasonable solution, depends on the portion of outliers in data. The computational time for BLAVE does not depend on the portion of outliers in the observations, but it grows with the number of observations, same as RANSAC.

A hybrid method combining artificial neural network (ANN) with genetic algorithm (GA) is discussed in this paper. A new strategy of optimization on ANN structure and weights based on multi-population GA is proposed, and the quantitative optimization of ANN is realized. The Levenberg-Marquardt(LM) algorithm is used for further training the neural network, which can avoid the weak local searching ability of GA and shows both of the merits of GA as well as ANN. In this paper, the algorithm proposed is employed in the electrical machine fault diagnosis, and the simulation results verified the correctness and effectiveness of the scheme proposed.