arXiv subjects

Taiping Zeng

Publications and source records attributed to Taiping Zeng.

15 recordsLinked to original sources

arXiv2026

WeaveLA: Event Driven Cross-Subtask Latent Memory Weaving for Repetitive Robot Manipulation

Vision-Language-Action (VLA) policies have achieved remarkable single-step manipulation, yet they remain brittle precisely where each stage depends on what was just completed. The core issue is structural: short-window VLAs lack an explicit channel for rouxting information across sub-task boundaries, and existing memory-augmented variants either write at every frame, retrieve from demonstration-time stages, or fire at sub-goal events without performing an explicit sub-task-to-sub-task hand-off into the action expert. We identify the sub-goal completion event as the natural temporal unit for cross-subtask memory hand-off, and present WeaveLA (Weave Latent memory for Vision-Language-Action policies), a cross-subtask memory interface that, on top of a frozen VLA backbone, compresses each completed segment into latent tokens via query-driven attention pooling and routes them directly into the action-generation path of the next sub-task. This event-triggered, action-side design preserves the base policy's short-window interface while adding a lightweight cross-subtask channel. Through stratified evaluation on RoboMME with a $\pi_{0.5}$ backbone, WeaveLA's gains land exactly where the channel is needed: on the hardest repetition slice (SwingXtimes, $N{=}3$), success rises from $0\%$ to $47.8\%$, while single-execution episodes remain unchanged. Per-episode paired analysis confirms the gains are confined to tasks whose causal structure requires cross-subtask information.

cs.CV↗

arXiv2026

MindVoice: Reconstructing Intelligible Speech from Non-invasive Neural Signals with Pretrained Priors

Reconstructing continuous speech from non-invasive neural recordings is a fundamental problem for probing human auditory perception and building safe, scalable speech brain-computer interfaces. Despite recent progress, intelligible reconstruction remains elusive, as non-invasive recordings are inherently noisy, spatially blurred, and only partially preserve information about perceived speech. Existing methods directly map neural activity to entangled speech representations before synthesizing waveforms with neural vocoders, resulting in spectral-similar but unintelligible results. To overcome these limitations, we introduce MindVoice, a neuro-to-speech reconstruction framework that uses pretrained models to compensate for the incomplete semantic and acoustic information in neural recordings. MindVoice disentangles reconstruction into two complementary pathways: one recovers high-level semantic content, while the other estimates fine-grained acoustic attributes. These inferred representations are then fused with powerful speech generation models and in-context voice cloning to synthesize natural and intelligible utterances. Extensive experiments on EEG and MEG demonstrate that MindVoice substantially outperforms existing methods on various metrics. These results show that pretrained priors provide a principled way to bridge the gap between noisy neural recordings and natural speech, highlighting a promising attempt for auditory neuroscience research and non-invasive speech brain-computer interfaces.

cs.SD↗

arXiv2025

DetAny4D: Detect Anything 4D Temporally in a Streaming RGB Video

Reliable 4D object detection, which refers to 3D object detection in streaming video, is crucial for perceiving and understanding the real world. Existing open-set 4D object detection methods typically make predictions on a frame-by-frame basis without modeling temporal consistency, or rely on complex multi-stage pipelines that are prone to error propagation across cascaded stages. Progress in this area has been hindered by the lack of large-scale datasets that capture continuous reliable 3D bounding box (b-box) annotations. To overcome these challenges, we first introduce DA4D, a large-scale 4D detection dataset containing over 280k sequences with high-quality b-box annotations collected under diverse conditions. Building on DA4D, we propose DetAny4D, an open-set end-to-end framework that predicts 3D b-boxes directly from sequential inputs. DetAny4D fuses multi-modal features from pre-trained foundational models and designs a geometry-aware spatiotemporal decoder to effectively capture both spatial and temporal dynamics. Furthermore, it adopts a multi-task learning architecture coupled with a dedicated training strategy to maintain global consistency across sequences of varying lengths. Extensive experiments show that DetAny4D achieves competitive detection accuracy and significantly improves temporal stability, effectively addressing long-standing issues of jitter and inconsistency in 4D object detection. Data and code will be released upon acceptance.

cs.CV↗

arXiv2025

EEG-fused Digital Twin Brain for Autonomous Driving in Virtual Scenarios

Current methodologies typically integrate biophysical brain models with functional magnetic resonance imaging(fMRI) data - while offering millimeter-scale spatial resolution (0.5-2 mm^3 voxels), these approaches suffer from limited temporal resolution (>0.5 Hz) for tracking rapid neural dynamics during continuous tasks. Conversely, Electroencephalogram (EEG) provides millisecond-scale temporal precision (<=1 ms sampling rate) for real-time guidance of continuous task execution, albeit constrained by low spatial resolution. To reconcile these complementary modalities, we present a generalizable Bayesian inference framework that integrates high-spatial-resolution structural MRI(sMRI) with high-temporal-resolution EEG to construct a biologically realistic digital twin brain(DTB) model. The framework establishes voxel-wise mappings between millisecond-scale EEG and sMRI-derived spiking networks, while demonstrating its translational potential through a brain-inspired autonomous driving simulation. Our EEG-DTB model achieves capabilities: (1) Biologically-plausible EEG signal generation (0.88 resting-state,0.60 task-state correlation), with simulated signals in task-state yielding steering predictions outperforming both chance and empirical signals (p<0.05); (2) Successful autonomous driving in the CARLA simulator using decoded steering angles. The proposed approach pioneers a new paradigm for studying sensorimotor integration and for mechanistic studies of perception-action cycles and the development of brain-inspired control systems.

q-bio.NC↗

arXiv2025

Dynamic Manifold Hopfield Networks for Context-Dependent Associative Memory

Neural population activity in cortical and hippocampal circuits can be flexibly reorganized by context, suggesting that cognition relies on dynamic manifolds rather than static representations. However, how such dynamic organization can be realized mechanistically within a unified dynamical system remains unclear. Continuous Hopfield networks provide a classical attractor framework in which neural dynamics follow gradient descent on a fixed energy landscape, constraining retrieval within a static attractor manifold geometry. Extending this approach, we introduce Dynamic Manifold Hopfield Networks (DMHN), continuous dynamical models in which contextual modulation dynamically reshapes attractor geometry, transforming a static attractor manifold into a context-dependent family of neural manifolds. In DMHN, network interactions are learned in a data-driven manner, to intrinsically deform the geometry of its attractor manifold across cues without explicit context-specific parameterization. As a result, in associative retrieval, DMHN achieve substantially higher capacity and robustness than classical and modern Hopfield networks: when storing $2N$ patterns in a network of $N$ neurons, DMHN attain reliable retrieval with an average accuracy of 64%, compared with 1% and 13% for classical and modern variants, respectively. Together, these results establish dynamic reorganization of attractor manifold geometry as a principled mechanism for context-dependent remapping in neural associative memory.

cs.LG↗

arXiv2025

Hi-Dyna Graph: Hierarchical Dynamic Scene Graph for Robotic Autonomy in Human-Centric Environments

Autonomous operation of service robotics in human-centric scenes remains challenging due to the need for understanding of changing environments and context-aware decision-making. While existing approaches like topological maps offer efficient spatial priors, they fail to model transient object relationships, whereas dense neural representations (e.g., NeRF) incur prohibitive computational costs. Inspired by the hierarchical scene representation and video scene graph generation works, we propose Hi-Dyna Graph, a hierarchical dynamic scene graph architecture that integrates persistent global layouts with localized dynamic semantics for embodied robotic autonomy. Our framework constructs a global topological graph from posed RGB-D inputs, encoding room-scale connectivity and large static objects (e.g., furniture), while environmental and egocentric cameras populate dynamic subgraphs with object position relations and human-object interaction patterns. A hybrid architecture is conducted by anchoring these subgraphs to the global topology using semantic and spatial constraints, enabling seamless updates as the environment evolves. An agent powered by large language models (LLMs) is employed to interpret the unified graph, infer latent task triggers, and generate executable instructions grounded in robotic affordances. We conduct complex experiments to demonstrate Hi-Dyna Grap's superior scene representation effectiveness. Real-world deployments validate the system's practicality with a mobile manipulator: robotics autonomously complete complex tasks with no further training or complex rewarding in a dynamic scene as cafeteria assistant. See https://anonymous.4open.science/r/Hi-Dyna-Graph-B326 for video demonstration and more details.

cs.RO↗

arXiv2025

LOG-Nav: Efficient Layout-Aware Object-Goal Navigation with Hierarchical Planning

We introduce LOG-Nav, an efficient layout-aware object-goal navigation approach designed for complex multi-room indoor environments. By planning hierarchically leveraging a global topologigal map with layout information and local imperative approach with detailed scene representation memory, LOG-Nav achieves both efficient and effective navigation. The process is managed by an LLM-powered agent, ensuring seamless effective planning and navigation, without the need for human interaction, complex rewards, or costly training. Our experimental results on the MP3D benchmark achieves 85\% object navigation success rate (SR) and 79\% success rate weighted by path length (SPL) (over 40\% point improvement in SR and 60\% improvement in SPL compared to exsisting methods). Furthermore, we validate the robustness of our approach through virtual agent and real-world robotic deployment, showcasing its capability in practical scenarios.

cs.RO↗

arXiv2025

Towards a Vision-Language Episodic Memory Framework: Large-scale Pretrained Model-Augmented Hippocampal Attractor Dynamics

Modeling episodic memory (EM) remains a significant challenge in both neuroscience and AI, with existing models either lacking interpretability or struggling with practical applications. This paper proposes the Vision-Language Episodic Memory (VLEM) framework to address these challenges by integrating large-scale pretrained models with hippocampal attractor dynamics. VLEM leverages the strong semantic understanding of pretrained models to transform sensory input into semantic embeddings as the neocortex, while the hippocampus supports stable memory storage and retrieval through attractor dynamics. In addition, VLEM incorporates prefrontal working memory and the entorhinal gateway, allowing interaction between the neocortex and the hippocampus. To facilitate real-world applications, we introduce EpiGibson, a 3D simulation platform for generating episodic memory data. Experimental results demonstrate the VLEM framework's ability to efficiently learn high-level temporal representations from sensory input, showcasing its robustness, interpretability, and applicability in real-world scenarios.

q-bio.NC↗

arXiv2024

Topo-Field: Topometric mapping with Brain-inspired Hierarchical Layout-Object-Position Fields

Mobile robots require comprehensive scene understanding to operate effectively in diverse environments, enriched with contextual information such as layouts, objects, and their relationships. Although advances like neural radiation fields (NeRFs) offer high-fidelity 3D reconstructions, they are computationally intensive and often lack efficient representations of traversable spaces essential for planning and navigation. In contrast, topological maps are computationally efficient but lack the semantic richness necessary for a more complete understanding of the environment. Inspired by a population code in the postrhinal cortex (POR) strongly tuned to spatial layouts over scene content rapidly forming a high-level cognitive map, this work introduces Topo-Field, a framework that integrates Layout-Object-Position (LOP) associations into a neural field and constructs a topometric map from this learned representation. LOP associations are modeled by explicitly encoding object and layout information, while a Large Foundation Model (LFM) technique allows for efficient training without extensive annotations. The topometric map is then constructed by querying the learned neural representation, offering both semantic richness and computational efficiency. Empirical evaluations in multi-room environments demonstrate the effectiveness of Topo-Field in tasks such as position attribute inference, query localization, and topometric planning, successfully bridging the gap between high-fidelity scene understanding and efficient robotic navigation.

cs.RO↗

arXiv2023

SUPS: A Simulated Underground Parking Scenario Dataset for Autonomous Driving

Automatic underground parking has attracted considerable attention as the scope of autonomous driving expands. The auto-vehicle is supposed to obtain the environmental information, track its location, and build a reliable map of the scenario. Mainstream solutions consist of well-trained neural networks and simultaneous localization and mapping (SLAM) methods, which need numerous carefully labeled images and multiple sensor estimations. However, there is a lack of underground parking scenario datasets with multiple sensors and well-labeled images that support both SLAM tasks and perception tasks, such as semantic segmentation and parking slot detection. In this paper, we present SUPS, a simulated dataset for underground automatic parking, which supports multiple tasks with multiple sensors and multiple semantic labels aligned with successive images according to timestamps. We intend to cover the defect of existing datasets with the variability of environments and the diversity and accessibility of sensors in the virtual scene. Specifically, the dataset records frames from four surrounding fisheye cameras, two forward pinhole cameras, a depth camera, and data from LiDAR, inertial measurement unit (IMU), GNSS. Pixel-level semantic labels are provided for objects, especially ground signs such as arrows, parking lines, lanes, and speed bumps. Perception, 3D reconstruction, depth estimation, and SLAM, and other relative tasks are supported by our dataset. We also evaluate the state-of-the-art SLAM algorithms and perception models on our dataset. Finally, we open source our virtual 3D scene built based on Unity Engine and release our dataset at https://github.com/jarvishou829/SUPS.

cs.CV↗

arXiv2020

StereoNeuroBayesSLAM: A Neurobiologically Inspired Stereo Visual SLAM System Based on Direct Sparse Method

We propose a neurobiologically inspired visual simultaneous localization and mapping (SLAM) system based on direction sparse method to real-time build cognitive maps of large-scale environments from a moving stereo camera. The core SLAM system mainly comprises a Bayesian attractor network, which utilizes neural responses of head direction (HD) cells in the hippocampus and grid cells in the medial entorhinal cortex (MEC) to represent the head direction and the position of the robot in the environment, respectively. Direct sparse method is employed to accurately and robustly estimate velocity information from a stereo camera. Input rotational and translational velocities are integrated by the HD cell and grid cell networks, respectively. We demonstrated our neurobiologically inspired stereo visual SLAM system on the KITTI odometry benchmark datasets. Our proposed SLAM system is robust to real-time build a coherent semi-metric topological map from a stereo camera. Qualitative evaluation on cognitive maps shows that our proposed neurobiologically inspired stereo visual SLAM system outperforms our previous brain-inspired algorithms and the neurobiologically inspired monocular visual SLAM system both in terms of tracking accuracy and robustness, which is closer to the traditional state-of-the-art one.

cs.RO↗

arXiv2019

Bayesian Integration of Multi-resolutional Grid Codes for Spatial Cognition

Fourier-like summation of several grid cell modules with different spatial frequencies in the medial entorhinal cortex (MEC) has long been proposed to form the contours of place firing fields. Recent experiments largely, but not completely, support this theory. Place fields are obviously expanded by inactivation of dorsal MEC, which fits the hypothesis. However, contrary to the prediction, inactivation of ventral MEC is also weakly broaden the spatial place firing patterns. In this study, we derive the model from grid spatial frequencies represented by Gaussian profiles to a 1D place field by Bayesian inference, and further provide completely theoretical explanations for expansion of place fields and predictions for alignments of grid components. To understand the information transform across between neocortex, entorhinal cortex, and hippocampus, we propose spatial memory indexing theory from hippocampal indexing theory to investigate how neural dynamics work in the entorhinal-hippocampal circuit. The inputs of place cells in CA3 are converged from three grid modules with different grid spacings layer II of MEC by Bayesian mechanism. We resort to the robot system to test Fourier hypothesis and spatial memory indexing theory, and validate our proposed entorhinal-hippocampal model. And then we demonstrate its cognitive mapping capability on the KITTI odometry benchmark dataset. Results suggest that our model provides a rational theoretical explanation for the biological experimental results. Results also show that the proposed model is robust for simultaneous localization and mapping (SLAM) in the large-scale environment. Our proposed model theoretically supports for Fourier hypothesis in a general Bayesian mechanism, which may pertain to other neural systems in addition to spatial cognition.

q-bio.NC↗

arXiv2019

Experience Dependent Formation of Global Coherent Representation of Environment by Grid Cells and Head Direction Cells

The grid firing patterns are thought to provide an efficient intrinsic metric capable of supporting universal spatial metric for mammalian spatial navigation in all environments. However, whether spatial representations of grid cells in the entorhinal cortex are determined by local environment cues or form globally coherent patterns remains undetermined. To explore this underlying mechanism, here we proposed a possible theoretical explanation to describe connection between the neural space and the physical environment and transition from a local anchored to a global coherent representation according to relationship between grid phase distance between physical distance in the physical environment, and tested our method based on simultaneous localization and mapping (SLAM) system on a iRat rodent-sized robot platform in a rat-like maze. Our robotic exploration experiments show that the grid firing firstly is determined by local environment cues, and after self-correction with experience-dependence, the regular, continuous grid firing patterns tessellate the explored space. Head direction (HD) cells also show global patterns in our experimental results. Our results support that grid firing patterns do provide a universal spatial metric for mammalian spatial navigation in complex environments. The results in this study also provide insights into experience-dependent interactions between path integrative calculation of location in the entorhinal cortex and learned associations to the external sensory cues in the hippocampus, which is likely to be critical understanding spatial memory, even episodic memory.

q-bio.NC↗

arXiv2019

Learning Sparse Spatial Codes for Cognitive Mapping Inspired by Entorhinal-Hippocampal Neurocircuit

The entorhinal-hippocampal circuit plays a critical role in higher brain functions, especially spatial cognition. Grid cells in the medial entorhinal cortex (MEC) periodically fire with different grid spacing and orientation, which makes a contribution that place cells in the hippocampus can uniquely encode locations in an environment. But how sparse firing granule cells in the dentate gyrus are formed from grid cells in the MEC remains to be determined. Recently, the fruit fly olfactory circuit provides a variant algorithm (called locality-sensitive hashing) to solve this problem. To investigate how the sparse place firing generates in the dentate gyrus can help animals to break the perception ambiguity during environment exploration, we build a biologically relevant, computational model from grid cells to place cells. The weight from grid cells to dentate gyrus granule cells is learned by competitive Hebbian learning. We resorted to the robot system for demonstrating our cognitive mapping model on the KITTI odometry benchmark dataset. The experimental results show that our model is able to stably, robustly build a coherent semi-metric topological map in the large-scale outdoor environment. The experimental results suggest that the entorhinal-hippocampal circuit as a variant locality-sensitive hashing algorithm is capable of generating sparse encoding for easily distinguishing different locations in the environment. Our experiments also provide theoretical supports that this analogous hashing algorithm may be a general principle of computation in different brain regions and species.

q-bio.NC↗

arXiv2019

A Brain-Inspired Compact Cognitive Mapping System

As the robot explores the environment, the map grows over time in the simultaneous localization and mapping (SLAM) system, especially for the large scale environment. The ever-growing map prevents long-term mapping. In this paper, we developed a compact cognitive mapping approach inspired by neurobiological experiments. Inspired from neighborhood cells, neighborhood fields determined by movement information, i.e. translation and rotation, are proposed to describe one of distinct segments of the explored environment. The vertices and edges with movement information below the threshold of the neighborhood fields are avoided adding to the cognitive map. The optimization of the cognitive map is formulated as a robust non-linear least squares problem, which can be efficiently solved by the fast open linear solvers as a general problem. According to the cognitive decision-making of familiar environments, loop closure edges are clustered depending on time intervals, and then parallel computing is applied to perform batch global optimization of the cognitive map for ensuring the efficiency of computation and real-time performance. After the loop closure process, scene integration is performed, in which revisited vertices are removed subsequently to further reduce the size of the cognitive map. A monocular visual SLAM system is developed to test our approach in a rat-like maze environment. Our results suggest that the method largely restricts the growth of the size of the cognitive map over time, and meanwhile, the compact cognitive map correctly represents the overall layer of the environment as the standard one. Experiments demonstrate that our method is very suited for compact cognitive mapping to support long-term robot mapping. Our approach is simple, but pragmatic and efficient for achieving the compact cognitive map.

cs.RO↗