arXiv subjects

Guanghui Ren

Publications and source records attributed to Guanghui Ren.

At least 37 records · Page 2Linked to original sources

arXiv2025

Act2Goal: From World Model To General Goal-conditioned Policy

Specifying robotic manipulation tasks in a manner that is both expressive and precise remains a central challenge. While visual goals provide a compact and unambiguous task specification, existing goal-conditioned policies often struggle with long-horizon manipulation due to their reliance on single-step action prediction without explicit modeling of task progress. We propose Act2Goal, a general goal-conditioned manipulation policy that integrates a goal-conditioned visual world model with multi-scale temporal control. Given a current observation and a target visual goal, the world model generates a plausible sequence of intermediate visual states that captures long-horizon structure. To translate this visual plan into robust execution, we introduce Multi-Scale Temporal Hashing (MSTH), which decomposes the imagined trajectory into dense proximal frames for fine-grained closed-loop control and sparse distal frames that anchor global task consistency. The policy couples these representations with motor control through end-to-end cross-attention, enabling coherent long-horizon behavior while remaining reactive to local disturbances. Act2Goal achieves strong zero-shot generalization to novel objects, spatial layouts, and environments. We further enable reward-free online adaptation through hindsight goal relabeling with LoRA-based finetuning, allowing rapid autonomous improvement without external supervision. Real-robot experiments demonstrate that Act2Goal improves success rates from 30% to 90% on challenging out-of-distribution tasks within minutes of autonomous interaction, validating that goal-conditioned world models with multi-scale temporal control provide structured guidance necessary for robust long-horizon manipulation. Project page: https://act2goal.github.io/

cs.RO↗

arXiv2025

High efficiency and compact lithium niobate non-resonant recirculating phase modulator and its applications

High modulation efficiency and a compact footprint are critical for next-generation electro-optic (EO) modulators. We introduce a new class of non-resonant recirculating phase modulators (PMs) that boosts modulation efficiency by repeatedly modulating the optical field within a single, non-resonant waveguide, while fundamentally removing the loop-length matching constraint that has limited prior recirculating schemes. This architectural breakthrough simultaneously enables a much smaller device footprint and an extended low-V$π$ bandwidth, without relying on narrowband resonances. Building on this concept, we experimentally demonstrate both a Mach-Zehnder modulator (MZM) and a cascaded PM, and verify their versatility in finite impulse response (FIR) filtering and optical frequency comb (OFC) generation. The recirculating MZM operates as a 4-tap rectangular-window FIR filter with 110 GHz bandwidth in a compact 2.889$\times$0.58 mm$^2$ footprint. The cascaded PM achieves a 3.40 GHz low-V$π$ bandwidth, a 110 GHz resonant EO bandwidth, and a V$π$L of 0.7 V$\cdot$cm, and generates 20 OFC lines under a 33 dBm microwave drive. These results demonstrate, for the first time, a practical and highly efficient non-resonant recirculating modulation platform, laying the groundwork for scalable high-order mode recirculating modulators (RMs) and opening new opportunities in optical communications, sensing, and microwave photonics.

physics.optics↗

arXiv2025

EnerVerse: Envisioning Embodied Future Space for Robotics Manipulation

We introduce EnerVerse, a generative robotics foundation model that constructs and interprets embodied spaces. EnerVerse employs a chunk-wise autoregressive video diffusion framework to predict future embodied spaces from instructions, enhanced by a sparse context memory for long-term reasoning. To model the 3D robotics world, we adopt a multi-view video representation, providing rich perspectives to address challenges like motion ambiguity and 3D grounding. Additionally, EnerVerse-D, a data engine pipeline combining generative modeling with 4D Gaussian Splatting, forms a self-reinforcing data loop to reduce the sim-to-real gap. Leveraging these innovations, EnerVerse translates 4D world representations into physical actions via a policy head (EnerVerse-A), achieving state-of-the-art performance in both simulation and real-world tasks. For efficiency, EnerVerse-A reuses features from the first denoising step and predicts action chunks, achieving about 280 ms per 8-step action chunk on a single RTX 4090. Further video demos, dataset samples could be found in our project page.

cs.RO↗

arXiv2025

Genie Envisioner: A Unified World Foundation Platform for Robotic Manipulation

We introduce Genie Envisioner (GE), a unified world foundation platform for robotic manipulation that integrates policy learning, evaluation, and simulation within a single video-generative framework. At its core, GE-Base is a large-scale, instruction-conditioned video diffusion model that captures the spatial, temporal, and semantic dynamics of real-world robotic interactions in a structured latent space. Built upon this foundation, GE-Act maps latent representations to executable action trajectories through a lightweight, flow-matching decoder, enabling precise and generalizable policy inference across diverse embodiments with minimal supervision. To support scalable evaluation and training, GE-Sim serves as an action-conditioned neural simulator, producing high-fidelity rollouts for closed-loop policy development. The platform is further equipped with EWMBench, a standardized benchmark suite measuring visual fidelity, physical consistency, and instruction-action alignment. Together, these components establish Genie Envisioner as a scalable and practical foundation for instruction-driven, general-purpose embodied intelligence. All code, models, and benchmarks will be released publicly.

cs.RO↗

arXiv2025

UniVLA: Learning to Act Anywhere with Task-centric Latent Actions

A generalist robot should perform effectively across various environments. However, most existing approaches heavily rely on scaling action-annotated data to enhance their capabilities. Consequently, they are often limited to single physical specification and struggle to learn transferable knowledge across different embodiments and environments. To confront these limitations, we propose UniVLA, a new framework for learning cross-embodiment vision-language-action (VLA) policies. Our key innovation is to derive task-centric action representations from videos with a latent action model. This enables us to exploit extensive data across a wide spectrum of embodiments and perspectives. To mitigate the effect of task-irrelevant dynamics, we incorporate language instructions and establish a latent action model within the DINO feature space. Learned from internet-scale videos, the generalist policy can be deployed to various robots through efficient latent action decoding. We obtain state-of-the-art results across multiple manipulation and navigation benchmarks, as well as real-robot deployments. UniVLA achieves superior performance over OpenVLA with less than 1/20 of pretraining compute and 1/10 of downstream data. Continuous performance improvements are observed as heterogeneous data, even including human videos, are incorporated into the training pipeline. The results underscore UniVLA's potential to facilitate scalable and efficient robot policy learning.

cs.RO↗

arXiv2025

Fidelity-Aware Data Composition for Robust Robot Generalization

Generalist robot policies trained on large-scale, visually homogeneous datasets can be susceptible to shortcut learning, which impairs their out-of-distribution (OOD) generalization. While generative data augmentation is a common approach to introduce diversity, it presents a subtle challenge: data composition. Naively mixing real and synthetic data can corrupt the learning signal, as this process often prioritizes visual diversity at the expense of information fidelity. This paper suggests that robust generalization depends on principled, fidelity-aware data composition. We introduce Coherent Information Fidelity Tuning (CIFT), a framework that treats data composition as an optimization problem. CIFT uses a practical proxy for Information Fidelity based on the feature-space geometry of a dataset. This enables the identification of a phase transition, termed the Decoherence Point, where training stability degrades. The framework includes a generative engine, Multi-View Video Augmentation (MVAug), to synthesize a causally disentangled data spectrum for this tuning process. Applying CIFT to policy architectures such as $π_0$ and Diffusion Policy improves OOD success rates by over 54\%. These results indicate that fidelity-aware composition, beyond data synthesis alone, is an important component for developing robust, general-purpose robots.

cs.RO↗

arXiv2025

AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems

We explore how scalable robot data can address real-world challenges for generalized robotic manipulation. Introducing AgiBot World, a large-scale platform comprising over 1 million trajectories across 217 tasks in five deployment scenarios, we achieve an order-of-magnitude increase in data scale compared to existing datasets. Accelerated by a standardized collection pipeline with human-in-the-loop verification, AgiBot World guarantees high-quality and diverse data distribution. It is extensible from grippers to dexterous hands and visuo-tactile sensors for fine-grained skill acquisition. Building on top of data, we introduce Genie Operator-1 (GO-1), a novel generalist policy that leverages latent action representations to maximize data utilization, demonstrating predictable performance scaling with increased data volume. Policies pre-trained on our dataset achieve an average performance improvement of 30% over those trained on Open X-Embodiment, both in in-domain and out-of-distribution scenarios. GO-1 exhibits exceptional capability in real-world dexterous and long-horizon tasks, achieving over 60% success rate on complex tasks and outperforming prior RDT approach by 32%. By open-sourcing the dataset, tools, and models, we aim to democratize access to large-scale, high-quality robot data, advancing the pursuit of scalable and general-purpose intelligence.

cs.RO↗

arXiv2025

Hume: Introducing System-2 Thinking in Visual-Language-Action Model

Humans practice slow thinking before performing actual actions when handling complex tasks in the physical world. This thinking paradigm, recently, has achieved remarkable advancement in boosting Large Language Models (LLMs) to solve complex tasks in digital domains. However, the potential of slow thinking remains largely unexplored for robotic foundation models interacting with the physical world. In this work, we propose Hume: a dual-system Vision-Language-Action (VLA) model with value-guided System-2 thinking and cascaded action denoising, exploring human-like thinking capabilities of Vision-Language-Action models for dexterous robot control. System 2 of Hume implements value-Guided thinking by extending a Vision-Language-Action Model backbone with a novel value-query head to estimate the state-action value of predicted actions. The value-guided thinking is conducted by repeat sampling multiple action candidates and selecting one according to state-action value. System 1 of Hume is a lightweight reactive visuomotor policy that takes System 2 selected action and performs cascaded action denoising for dexterous robot control. At deployment time, System 2 performs value-guided thinking at a low frequency while System 1 asynchronously receives the System 2 selected action candidate and predicts fluid actions in real time. We show that Hume outperforms the existing state-of-the-art Vision-Language-Action models across multiple simulation benchmark and real-robot deployments.

cs.RO↗

arXiv2025

EWMBench: Evaluating Scene, Motion, and Semantic Quality in Embodied World Models

Recent advances in creative AI have enabled the synthesis of high-fidelity images and videos conditioned on language instructions. Building on these developments, text-to-video diffusion models have evolved into embodied world models (EWMs) capable of generating physically plausible scenes from language commands, effectively bridging vision and action in embodied AI applications. This work addresses the critical challenge of evaluating EWMs beyond general perceptual metrics to ensure the generation of physically grounded and action-consistent behaviors. We propose the Embodied World Model Benchmark (EWMBench), a dedicated framework designed to evaluate EWMs based on three key aspects: visual scene consistency, motion correctness, and semantic alignment. Our approach leverages a meticulously curated dataset encompassing diverse scenes and motion patterns, alongside a comprehensive multi-dimensional evaluation toolkit, to assess and compare candidate models. The proposed benchmark not only identifies the limitations of existing video generation models in meeting the unique requirements of embodied tasks but also provides valuable insights to guide future advancements in the field. The dataset and evaluation tools are publicly available at https://github.com/AgibotTech/EWMBench.

cs.RO↗

arXiv2025

Single Vanadium Dioxide Nanoparticle-Enabled Plasmonic switch with Thermal and Electronic Reconfigurability

We present an integrated switch that combines plasmonic and neuromorphic technologies with a single sub-stoichiometric VO2-x nanoparticle. The presented device acts as a versatile plasmonic switch with dual thermal and electrical reconfigurability leveraging the near-room temperature phase transition of the VO2-x nanoparticles combined with the rapid phase recovery to drive the device. The change in both the optical and electrical properties of the VO2-x nanoparticle enables simultaneous optical and electrical readouts making the plasmonic device suitable as a phase change memory cell which is crucial in the convergence of computing and communication technologies. Our demonstration of reversible electrical switching, evidenced by a 6dB modulation depth and concurrent optical and electrical outputs, signifies a major stride in merging electronic and photonic functionalities within phase-change material devices. This novel strategy not only harmonizes optical communication with electronic computing but also advances the development of sophisticated integrated neuromorphic devices.

physics.optics↗

arXiv2025

EnerVerse-AC: Envisioning Embodied Environments with Action Condition

Robotic imitation learning has advanced from solving static tasks to addressing dynamic interaction scenarios, but testing and evaluation remain costly and challenging due to the need for real-time interaction with dynamic environments. We propose EnerVerse-AC (EVAC), an action-conditional world model that generates future visual observations based on an agent's predicted actions, enabling realistic and controllable robotic inference. Building on prior architectures, EVAC introduces a multi-level action-conditioning mechanism and ray map encoding for dynamic multi-view image generation while expanding training data with diverse failure trajectories to improve generalization. As both a data engine and evaluator, EVAC augments human-collected trajectories into diverse datasets and generates realistic, action-conditioned video observations for policy testing, eliminating the need for physical robots or complex simulations. This approach significantly reduces costs while maintaining high fidelity in robotic manipulation evaluation. Extensive experiments validate the effectiveness of our method. Code, checkpoints, and datasets can be found at .

cs.RO↗

arXiv2025

Suspended Z-cut lithium niobate waveguides for stimulated Brillouin scattering

On-chip stimulated Brillouin scattering (SBS) has recently been demonstrated in thin-film lithium niobate (TFLN), an emerging material platform for integrated photonics offering large electro-optic and nonlinear properties. While previous work on SBS in TFLN have focused on surface SBS, in this contribution we experimentally demonstrate, for the first time, backward intra-modal SBS generation in suspended Z-cut TFLN waveguides. Our results show trapping of multiple acoustic modes in this structure, featuring a multi-peak Brillouin gain spectrum due to the excitation of higher-order acoustic modes. The findings expand the TFLN waveguide platform exploration for SBS interactions and provide a crucial step towards realizing optical processors for microwave signals or sensors integrated on TFLN.

physics.optics↗

arXiv2025

On-chip GaAs carrier lifetime investigation

We propose an alternative way to determine GaAs carrier lifetime using pump-probe measurement based on fibre optics and integrated waveguides. We find that our GaAs samples have the lifetime ranging from 30-80 ps, supporting the bandwidth $\geq$ 12.5 GHz. The platform utilised in this work could offer a cost-effective way to investigate photocarrier lifetime. Moreover, it may have the potential for high-speed switches and detectors, or could be exploited for possible all-optical modulation.

physics.optics↗

arXiv2025

Adversarial Data Collection: Human-Collaborative Perturbations for Efficient and Robust Robotic Imitation Learning

The pursuit of data efficiency, where quality outweighs quantity, has emerged as a cornerstone in robotic manipulation, especially given the high costs associated with real-world data collection. We propose that maximizing the informational density of individual demonstrations can dramatically reduce reliance on large-scale datasets while improving task performance. To this end, we introduce Adversarial Data Collection, a Human-in-the-Loop (HiL) framework that redefines robotic data acquisition through real-time, bidirectional human-environment interactions. Unlike conventional pipelines that passively record static demonstrations, ADC adopts a collaborative perturbation paradigm: during a single episode, an adversarial operator dynamically alters object states, environmental conditions, and linguistic commands, while the tele-operator adaptively adjusts actions to overcome these evolving challenges. This process compresses diverse failure-recovery behaviors, compositional task variations, and environmental perturbations into minimal demonstrations. Our experiments demonstrate that ADC-trained models achieve superior compositional generalization to unseen task instructions, enhanced robustness to perceptual perturbations, and emergent error recovery capabilities. Strikingly, models trained with merely 20% of the demonstration volume collected through ADC significantly outperform traditional approaches using full datasets. These advances bridge the gap between data-centric learning paradigms and practical robotic deployment, demonstrating that strategic data acquisition, not merely post-hoc processing, is critical for scalable, real-world robot learning. Additionally, we are curating a large-scale ADC-Robotics dataset comprising real-world manipulation tasks with adversarial perturbations. This benchmark will be open-sourced to facilitate advancements in robotic imitation learning.

cs.RO↗

arXiv2024

One Million Quality Factor Integrated Ring Resonators in the Mid-Infrared

We report ring resonators on a silicon germanium on silicon platform operating in the mid-infrared wavelength range around 3.5 - 4.6 μm with quality factors reaching up to one million. Advances in fabrication technology enable us to demonstrate such high Q-factors, which put silicon germanium at the forefront of mid-infrared integrated photonic platforms. The achievement of high Q is attested by the observation of degeneracy lifting between clockwise (CW) and counter-clockwise (CCW) resonances, as well as optical bistability due to an efficient power buildup in the rings.

physics.optics↗

arXiv2024

Ultra-wideband integrated microwave photonic multi-parameter measurement system on thin-film lithium niobate

Research on microwave signal measurement techniques is risen, driven by the expanding urgent demands of wireless communication, global positioning systems, remote sensing and 6G networks. In stark contrast with traditional electronic-based realization, the implementations of microwave signal measurement systems based on integrated compact photonic chip have exhibited distinct advantages in high operation bandwidth, light weight, and strong immunity to electromagnetic interference. However, although numerous integrated microwave photonic signal measurement systems have been reported, measurement bandwidth of the majority of them is still below 30 GHz due to the bandwidth limitation of electro-optical modulators (EOMs). Furthermore, previous studies often are more focused on the measurement of one single parameter (typically the frequency) of microwave signals, which has hindered their practical application in complex situations. Here, an integrated photonic microwave multi-parameter measurement system composed of microwave frequency measurement module and microwave phase amplitude measurement module based on thin-film lithium niobate (TFLN) platform is reported. Utilizing this system, not only the ultra-high bandwidth (up to 60GHz) of microwave frequency, phase and amplitude measurement with low root-mean-squares errors (450MHz, 3.43° and 1.64% of the measurement for frequency, phase and amplitude, respectively), but also the time-domain reconstruction of sinusoidal microwave signals is achieved. This demonstration further broadens the application of integrated TFLN photonic devices in microwave signal measurement technology to address the bandwidth bottleneck of the ever-growing microwave networks in the future information society.

physics.optics↗

arXiv2023

Maximizing the performance for microcomb based microwave photonic transversal signal processors

Microwave photonic (MWP) transversal signal processors offer a compelling solution for realizing versatile high-speed information processing by combining the advantages of reconfigurable electrical digital signal processing and high-bandwidth photonic processing. With the capability of generating a number of discrete wavelengths from micro-scale resonators, optical microcombs are powerful multi-wavelength sources for implementing MWP transversal signal processors with significantly reduced size, power consumption, and complexity. By using microcomb-based MWP transversal signal processors, a diverse range of signal processing functions have been demonstrated recently. In this paper, we provide a detailed analysis for the processing inaccuracy that is induced by the imperfect response of experimental components. First, we investigate the errors arising from different sources including imperfections in the microcombs, the chirp of electro-optic modulators, chromatic dispersion of the dispersive module, shaping errors of the optical spectral shapers, and noise of the photodetector. Next, we provide a global picture quantifying the impact of different error sources on the overall system performance. Finally, we introduce feedback control to compensate the errors caused by experimental imperfections and achieve significantly improved accuracy. These results provide a guide for optimizing the accuracy of microcomb-based MWP transversal signal processors.

eess.SP↗

arXiv2023

Hitless memory-reconfigurable photonic reservoir computing architecture

Reservoir computing is an analog bio-inspired computation model for efficiently processing time-dependent signals, the photonic implementations of which promise a combination of massive parallel information processing, low power consumption, and high speed operation. However, most implementations, especially for the case of time-delay reservoir computing (TDRC), require signal attenuation in the reservoir to achieve the desired system dynamics for a specific task, often resulting in large amounts of power being coupled outside of the system. We propose a novel TDRC architecture based on an asymmetric Mach-Zehnder interferometer (MZI) integrated in a resonant cavity which allows the memory capacity of the system to be tuned without the need for an optical attenuator block. Furthermore, this can be leveraged to find the optimal value for the specific components of the total memory capacity metric. We demonstrate this approach on the temporal bitwise XOR task and conclude that this way of memory capacity reconfiguration allows optimal performance to be achieved for memory-specific tasks.

eess.SP↗