arXiv subjects

Xinhong Zhang

Publications and source records attributed to Xinhong Zhang.

6 recordsLinked to original sources

arXiv2026

FlowPilot: Real-Time World-Action Modeling for Agile UAV Navigation

We present FlowPilot, a compact world-action model for real-time onboard UAV navigation from depth. Unlike map-then-optimize pipelines that require local reconstruction or end-to-end policies that lack explicit scene prediction, FlowPilot jointly denoises future depth observations and executable trajectories with flow matching. A dual-stream mixture-of-transformers couples video and action experts through shared attention, allowing future-scene prediction and trajectory generation to inform each other. At deployment, the model runs action-centrically and outputs only a trajectory. To ensure trackability, actions are parameterized as degree-7 Bernstein polynomials: the current state constrains the initial control points, and the network predicts five free control points, yielding C^2-continuous references with closed-form velocity, acceleration and jerk. FlowPilot is trained on a three-level depth pyramid spanning high-throughput simulation, photorealistic simulation, and real onboard data. In closed-loop simulation, it outperforms learning- and optimization-based baselines under increasing clutter and commanded speeds up to 8m/s. On a physical quadrotor, the full perception-to-action pipeline runs in under 18ms on a Jetson Orin NX and reaches 5.5m/s in cluttered indoor and forest environments using only onboard sensing and computation.

cs.RO↗

arXiv2026

AeroAct: Action-Centered World-Action Models for Language-Conditioned Quadrotor Flight

Language-conditioned quadrotor flight requires a policy to ground semantic goals, anticipate the visual consequences of ego-motion, and output control references that remain smooth and dynamically executable under rapidly changing first-person views. Existing aerial vision-language navigation and vision-language-action methods commonly use discrete actions, high-level waypoints, or instantaneous velocity commands, which provide limited supervision about how flight actions change future observations. We present AeroAct, an action-centered world-action model (WAM) for quadrotor navigation. To the best of our knowledge, AeroAct is the first WAM instantiated and demonstrated for real-world aerial flight. The model adapts a pretrained video diffusion Transformer to predict local trajectory-action chunks from egocentric visual history, proprioception, and language. Future first-person frames are used during training as dense consequence supervision, while deployment directly decodes actions without generating future video. To obtain aligned visual, state, language, and dynamically feasible action data, we build a DiffAero-based pipeline with complementary Isaac Lab and 3D Gaussian splatting renderers. We further introduce a low-cost handheld collection device that couples camera observations with motion estimates to recreate flight-like egocentric trajectories, and a self-guidance procedure that improves temporal consistency across overlapping trajectory chunks. Closed-loop simulation and real-world experiments show that temporal visual context improves target tracking and object-search performance, and that WAM-based policies can be executed on a physical quadrotor.

cs.RO↗

arXiv2026

MAD: Mapping-Aware World Models for Agile Quadrotor Flight

Agile quadrotor flight in cluttered scenes requires more than a reactive mapping from a depth image to a control command: the vehicle must remember which regions have been observed, infer nearby occupied space, and act under partial visibility and tight latency. In this paper, we present Mapping-Aware Dreamer (MAD), a geometry-aware world model for vision-based quadrotor flight. Instead of using raw-image reconstruction as the main self-supervised objective, MAD learns recurrent latent dynamics that reconstruct robocentric occupancy and visibility grid maps together with proprioceptive states. This design forces the latent state to encode local geometry, visibility history, and ego-motion in a form that is directly relevant to collision avoidance. MAD is trained in DiffAero using a GPU-parallel map-construction module that provides high-throughput supervision for occupancy and visibility. The learned representation is used in three policy-learning modes: imagination-based MAD-Dreamer and feature-extractor variants based on PPO and SHAC. Across visual navigation and racing tasks, MAD-based agents achieve higher success rates, faster flight, and better cross-task transfer than corresponding vision-only baselines. The model also produces interpretable map predictions and accurate ego-motion estimates from depth observations. We further deploy the learned policy on a physical quadrotor with an Intel RealSense D435i and demonstrate safe indoor and outdoor flight under limited sensing, reaching 9.66 m/s in simulation and 5.05 m/s in real-world forest experiments. These results show that mapping-aware world models provide a practical middle ground between modular aerial navigation and end-to-end learning.

cs.RO↗

arXiv2025

DiffAero: A GPU-Accelerated Differentiable Simulation Framework for Efficient Quadrotor Policy Learning

This letter introduces DiffAero, a lightweight, GPU-accelerated, and fully differentiable simulation framework designed for efficient quadrotor control policy learning. DiffAero supports both environment-level and agent-level parallelism and integrates multiple dynamics models, customizable sensor stacks (IMU, depth camera, and LiDAR), and diverse flight tasks within a unified, GPU-native training interface. By fully parallelizing both physics and rendering on the GPU, DiffAero eliminates CPU-GPU data transfer bottlenecks and delivers orders-of-magnitude improvements in simulation throughput. In contrast to existing simulators, DiffAero not only provides high-performance simulation but also serves as a research platform for exploring differentiable and hybrid learning algorithms. Extensive benchmarks and real-world flight experiments demonstrate that DiffAero and hybrid learning algorithms combined can learn robust flight policies in hours on consumer-grade hardware. The code is available at https://github.com/flyingbitac/diffaero.

cs.RO↗

arXiv2018

The (1,2)-step competition graph of a hypertournament

Competition graphs were created in connected to a biological model as a means of reflecting the competition relations among the predators in the food webs and determining the smallest dimension of ecological phase space. In 2011, Factor and Merz introduced the (1,2)-step competition graph of a digraph. Given a digraph $D=(V,A)$, the (1,2)-step competition graph of $D$, denoted $C_{1,2}(D)$, is a graph on $V(D)$ where $xy\in E(C_{1,2}(D))$ if and only if there exists a vertex $z\neq x,y$ such that either $d_{D-y}(x,z)=1$ and $d_{D-x}(y,z)\leq 2$ or $d_{D-x}(y,z)=1$ and $d_{D-y}(x,z)\leq 2$. They also characterized the (1,2)-step competition graphs of tournaments and extended some results to the $(i,j)$-step competition graphs of tournaments. In this paper, the definition of the (1,2)-step competition graph of a digraph is generalized to the one of a hypertournament and the $(1,2)$-step competition graph of a $k$-hypertournament is characterized. Also, the results are extended to the $(i,j)$-step competition graph of a $k$-hypertournament.

math.CO↗

arXiv2018

The H-force sets of the graphs satisfying the condition of Ore's theorem

Let $G$ be a Hamiltonian graph with $n$ vertices. A nonempty vertex set $X\subseteq V(G)$ is called a Hamiltonian cycle enforcing set (in short, an $H$-force set) of $G$ if every $X$-cycle of $G$ (i.e., a cycle of $G$ containing all vertices of $X$) is a Hamiltonian cycle. For the graph $G$, $h(G)$ is the smallest cardinality of an $H$-force set of $G$ and call it the $H$-force number of $G$. Ore's theorem states that the graph $G$ is Hamiltonian if $d(u)+d(v)\geq n$ for every pair of nonadjacent vertices $u,v$ of $G$. In this paper, we study the $H$-force sets of the graphs satisfying the condition of Ore's theorem, show that the $H$-force number of these graphs is possibly $n$, or $n-2$, or $\frac{n}{2}$ and give a classification of these graphs due to the $H$-force number.

math.CO↗