arXiv subjects

Ruogu Li

Publications and source records attributed to Ruogu Li.

7 recordsLinked to original sources

arXiv2026

MAC-I$^2$: Learned Metrics-Aware Covariance for Robust Visual-Inertial Fusion in Initialization and Calibration

Visual-Inertial (VI) fusion is fundamental to accurate and robust state estimation, where camera and IMU measurements are combined according to their respective uncertainties. Existing methods, however, fuse the two modalities with predefined uncertainties, regardless of how reliable each is in the local context, and thus often struggle under challenging environments involving illumination changes, dynamic objects, and textureless regions. In this paper, we present MAC-I$^2$, which achieves robust VI fusion through learned metric-aware covariance for both modalities, so that vision and IMU compete on their own merits rather than relying on predefined uncertainties. Here, metrics-aware means that each predicted covariance faithfully reflects the actual magnitude of the corresponding measurement noise. On the visual side, we propagate learned feature-matching uncertainties into pose covariances for the fusion. On the inertial side, motivated by the observation that integration error accumulates sharply at the early stage and grows slowly afterward, we design a learned IMU model with a learnable initial covariance, and propose a dedicated fine-tuning strategy on a held-out training subset to enable the metrics-aware covariance on unseen sequences. As a showcase, we build a VI initialization and calibration system, since accurate and robust initialization and calibration are the prerequisite for any reliable VI system. Experiments on EuRoC, and VBR show that MAC-I$^2$ substantially outperforms existing methods: it achieves a 99.9% initialization success rate on EuRoC, reducing gravity and velocity errors by about 60% and 42% over the strongest baseline, and maintains 80% success rate on challenging VBR sequences where baseline methods such as VINS-Mono drop below 10%.

cs.RO↗

arXiv2026

Handroid: Bridging Dexterous Hand and Humanoid

Dexterous hands and humanoid robots are typically developed as distinct embodiments: the former enable contact-rich manipulation at the object scale, whereas the latter provide mobility and whole-body interaction in human-centered environments. We introduce \textbf{Handroid}, a desktop-scale dual-embodiment robot that integrates both capabilities within a single reconfigurable platform. Handroid reuses one 27-DoF electromechanical body as either a dexterous hand or a desktop humanoid, measuring 0.33 m in height and 2.05 kg in weight. In the dexterous hand embodiment, 20 DoFs form an anthropomorphic hand closely matching the kinematic structure of the human hand. In the humanoid embodiment, the same articulated modules are reconfigured into a humanoid with a head, arms, and legs, including a 12-DoF lower-limb structure for locomotion and whole-body motion. Handroid further provides a unified control and learning framework supporting hand teleoperation, dexterous grasping, in-hand manipulation, humanoid locomotion, gait generation, and interactive motion authoring. We validate the platform through real-world dexterous manipulation, reinforcement-learning-based locomotion, keyframe motion deployment, and a long-horizon task involving embodiment reconfiguration, locomotion, docking, and dexterous pick-and-place. These results position Handroid as a compact and reproducible platform for advancing morphology-reconfigurable robotics and cross-embodiment robot learning.

cs.RO↗

arXiv2026

Current as Touch: Proprioceptive Contact Feedback for Compliant Dexterous Manipulation

Compliance is essential for dexterous manipulation, yet existing solutions often rely on external tactile or force sensors that are costly, fragile, and difficult to deploy on low-cost robot hands. We propose a proprioception-driven framework that learns contact-aware compliance cues from motor current and joint states. Since motor current is closely related to actuator torque, it provides an intrinsic signal for perceiving contact force, object resistance, and grasp stability without additional sensing hardware. Rather than estimating external wrenches or commanding torque, our method predicts a compliance reference position: an ideal joint-position target for a standard PD controller whose induced position error generates appropriate grasping force. This position-based formulation is compatible with mainstream teleoperation and policy-learning pipelines, while enabling the robot to adapt interaction forces from real-time proprioceptive feedback. Thus, motor current serves not only as a force proxy but also as a learnable proprioceptive contact signal for compliance reference prediction. Experiments on multiple dexterous hands and contact-rich tasks, including fragile object handling, sustained surface contact, thin-object retrieval, and dynamic load adaptation, show stable compliant grasping, safer and more efficient teleoperation, and improved downstream policy learning without external tactile or force sensors.

cs.RO↗

arXiv2026

SldprtNet: A Large-Scale Multimodal Dataset for CAD Generation in Language-Driven 3D Design

We introduce SldprtNet, a large-scale dataset comprising over 242,000 industrial parts, designed for semantic-driven CAD modeling, geometric deep learning, and the training and fine-tuning of multimodal models for 3D design. The dataset provides 3D models in both .step and .sldprt formats to support diverse training and testing. To enable parametric modeling and facilitate dataset scalability, we developed supporting tools, an encoder and a decoder, which support 13 types of CAD commands and enable lossless transformation between 3D models and a structured text representation. Additionally, each sample is paired with a composite image created by merging seven rendered views from different viewpoints of the 3D model, effectively reducing input token length and accelerating inference. By combining this image with the parameterized text output from the encoder, we employ the lightweight multimodal language model Qwen2.5-VL-7B to generate a natural language description of each part's appearance and functionality. To ensure accuracy, we manually verified and aligned the generated descriptions, rendered images, and 3D models. These descriptions, along with the parameterized modeling scripts, rendered images, and 3D model files, are fully aligned to construct SldprtNet. To assess its effectiveness, we fine-tuned baseline models on a dataset subset, comparing image-plus-text inputs with text-only inputs. Results confirm the necessity and value of multimodal datasets for CAD generation. It features carefully selected real-world industrial parts, supporting tools for scalable dataset expansion, diverse modalities, and ensured diversity in model complexity and geometric features, making it a comprehensive multimodal dataset built for semantic-driven CAD modeling and cross-modal learning.

cs.RO↗

arXiv2025

Co-Me: Confidence-Guided Token Merging for Visual Geometric Transformers

We propose Confidence-Guided Token Merging (Co-Me), an acceleration mechanism for visual geometric transformers without retraining or finetuning the base model. Co-Me distilled a light-weight confidence predictor to rank tokens by uncertainty and selectively merge low-confidence ones, effectively reducing computation while maintaining spatial coverage. Compared to similarity-based merging or pruning, the confidence signal in Co-Me reliably indicates regions emphasized by the transformer, enabling substantial acceleration without degrading performance. Co-Me applies seamlessly to various multi-view and streaming visual geometric transformers, achieving speedups that scale with sequence length. When applied to VGGT and Pi3, Co-Me achieves up to 21.5x and 20.4x speedup, making visual geometric transformers practical for real-time 3D perception and reconstruction.

cs.CV↗

arXiv2014

Throughput-Optimal Scheduling Design with Regular Service Guarantees in Wireless Networks

Motivated by the regular service requirements of video applications for improving Quality-of-Experience (QoE) of users, we consider the design of scheduling strategies in multi-hop wireless networks that not only maximize system throughput but also provide regular inter-service times for all links. Since the service regularity of links is related to the higher-order statistics of the arrival process and the policy operation, it is highly challenging to characterize and analyze directly. We overcome this obstacle by introducing a new quantity, namely the time-since-last-service (TSLS), which tracks the time since the last service. By combining it with the queue-length in the weight, we propose a novel maximum-weight type scheduling policy, called Regular Service Guarantee (RSG) Algorithm. The unique evolution of the TSLS counter poses significant challenges for the analysis of the RSG Algorithm. To tackle these challenges, we first propose a novel Lyapunov function to show the throughput optimality of the RSG Algorithm. Then, we prove that the RSG Algorithm can provide service regularity guarantees by using the Lyapunov-drift based analysis of the steady-state behavior of the stochastic processes. In particular, our algorithm can achieve a degree of service regularity within a factor of a fundamental lower bound we derive. This factor is a function of the system statistics and design parameters and can be as low as two in some special networks. Our results, both analytical and numerical, exhibit significant service regularity improvements over the traditional throughput-optimal policies, which reveals the importance of incorporating the metric of time-since-last-service into the scheduling policy for providing regulated service.

cs.NI↗

arXiv2010

Delay-Aware Cross-Layer Design for Network Utility Maximization in Multi-hop Networks

We investigate the problem of designing delay-aware joint flow control, routing, and scheduling algorithms in general multi-hop networks for maximizing network utilization. Since the end-to-end delay performance has a complex dependence on the high-order statistics of cross-layer algorithms, earlier optimization-based design methodologies that optimize the long term network utilization are not immediately well-suited for delay-aware design. This motivates us in this work to develop a novel design framework and alternative methods that take advantage of several unexploited design choices in the routing and the scheduling strategy spaces. In particular, we reveal and exploit a crucial characteristic of back pressure-type controllers that enables us to develop a novel link rate allocation strategy that not only optimizes long-term network utilization, but also yields loop free multi-path routes} between each source-destination pair. Moreover, we propose a regulated scheduling strategy, based on a token-based service discipline, for shaping the per-hop delay distribution to obtain highly desirable end-to-end delay performance. We establish that our joint flow control, routing, and scheduling algorithm achieves loop-free routes and optimal network utilization. Our extensive numerical studies support our theoretical results, and further show that our joint design leads to substantial end-to-end delay performance improvements in multi-hop networks compared to earlier solutions.

math.OC↗