The development of general robotic systems capable of manipulating in unstructured environments is a significant challenge. While Vision-Language Models(VLM) excel in high-level commonsense reasoning, they lack the fine-grained 3D spatial understanding required for precise manipulation tasks. Fine-tuning VLM on robotic datasets to create Vision-Language-Action Models(VLA) is a potential solution, but it is hindered by high data collection costs and generalization issues. To address these challenges, we propose a novel object-centric representation that bridges the gap between VLM’s high-level reasoning and the low-level precision required for manipulation. Our key insight is that an object’s canonical space, defined by its functional affordances, provides a structured and semantically meaningful way to describe interaction primitives, such as points and directions. These primitives act as a bridge, translating VLM’s commonsense reasoning into actionable 3D spatial constraints. In this context, we introduce a dual closed-loop, open-vocabulary robotic manipulation system: one loop for high-level planning through primitive resampling, interaction rendering and VLM checking, and another for low-level execution via 6D pose tracking. This design ensures robust, real-time control without requiring VLM fine-tuning. Extensive experiments demonstrate strong zero-shot generalization across diverse robotic manipulation tasks, highlighting the potential of this approach for automating large-scale data generation.

Over the past few decades, electromagnetic field signals have been widely used to develop data interpretation methods for the ground controlled source electromagnetic (CSEM) method in the frequency domain. However, identifying anomalous features from these signals remains challenging, as the intensity of the signals linked to these anomalies is weaker. To address this issue, this paper presents an inversion algorithm that utilizes apparent resistivity data to enhance anomaly detection. Due to the high computational complexity and memory demands of 3D CSEM inversion modeling, many real-world surveys are interpreted using 1D inversion schemes, which simplifies the modeling process but may limit accuracy. A quasi-2D inversion imaging technique based on the tracking balance operator (hereafter referred to as QITB) is introduced to balance efficiency and interpretive accuracy. Rather than performing 3D inversion, QITB applies 1D modeling in a structured framework to approximate 3D subsurface structures. This imaging-focused algorithm eliminates the need for complex 3D computations while maintaining high accuracy in identifying subsurface anomalies. Numerical simulations demonstrate that wide-field apparent resistivity is significantly more effective than electric field data in highlighting subsurface anomalies. Inversion results from synthetic data show that QITB identifies anomalies and yields more accurate estimates of burial depth. In contrast, conventional 1D inversion without the tracking balance operator, although capable of fitting 3D data, fails to approximate the true 3D model closely. In conclusion, QITB is an innovative and efficient method for rapidly interpreting 3D ground CSEM data, combining computational efficiency with reliable subsurface structure identification.

In the application of Bayesian networks to solve practical problems, it is likely to encounter the situation that the data set is expensive and difficult to obtain in large quantities and the small data set is easy to cause the inaccuracy of Bayesian network (BN) scoring functions, which affects the BN optimization results. Therefore, how to better learn Bayesian network structures under a small data set is an important problem we need to pay attention to and solve. This paper introduces the idea of parallel ensemble learning and proposes a new hybrid Bayesian network structure learning algorithm. The algorithm adopts the elite-based structure learner using genetic algorithm (ESL-GA) as the base learner. Firstly, the adjacency matrices of the network structures learned by ESL-GA are weighted and averaged. Then, according to the preset threshold, the edges between variables with weak dependence are filtered to obtain a fusion matrix. Finally, the fusion matrix is modified as the adjacency matrix of the integrated Bayesian network so as to obtain the final Bayesian network structure. Comparative experiments on the standard Bayesian network data sets show that the accuracy and reliability of the proposed algorithm are significantly better than other algorithms.