arXiv subjects

Yuechen Wu

Publications and source records attributed to Yuechen Wu.

3 recordsLinked to original sources

arXiv2026

SoftNav: Injecting 3D Scene Tokens into VLMs for Embodied Navigation

In goal-directed embodied navigation, where an agent must locate a specified target in an unseen environment, 3D scene understanding and navigation reasoning must work in concert. Current approaches transmit 3D scene information to vision-language models (VLMs) through text, suggesting a representation gap in our tested configurations; a controlled ablation confirms that direct embedding-level transfer significantly outperforms the evaluated text serialization formats. We introduce SoftNav, which injects entity-level 3D continuous representations -- one token per detected object or frontier -- into a VLM's hidden space as soft tokens through a lightweight projector. With the 3D encoder and VLM frozen, only ~1,200 samples and ~17M trainable parameters are needed. On HM3D-OVON, SoftNav achieves 74.2%/68.3%/66.7% SR across three splits, surpassing all prior methods in both SR and SPL; the same navigation policy transfers zero-shot to GOAT-Bench (67.2% SR), SG3D (47.2% s-SR), and real-world robot deployment without retraining or architectural modification. Injecting 3D scene tokens directly into VLMs bridges the representation gap, enabling transferable navigation with minimal training.

cs.RO↗

arXiv2026

NavCMPO: Critic-Guided MeanFlow Policy Optimization for Adaptive Navigation

End-to-end diffusion-based policies have demonstrated strong performance in mapless visual navigation, but their iterative denoising process introduces substantial inference latency, while behavior cloning limits performance to the quality of expert demonstrations. We present NavCMPO, a two-stage adaptive navigation framework that combines few-step MeanFlow trajectory generation, critic-guided refinement, and reinforcement learning fine-tuning. During pre-training, an obstacle proximity prediction task encourages the visual representation to capture obstacle-aware spatial information. To compensate for the degradation in obstacle avoidance caused by few-step generation, Critic-Guided Trajectory Refinement (CGTR) uses gradients from a critic trained with obstacle-point-cloud supervision to refine intermediate trajectories. During adaptation, the MeanFlow policy is fine-tuned using Proximal Policy Optimization with behavior-cloning regularization, while the critic is updated to accommodate embodiment-specific observation changes. Under a matched training budget on the InternVLA-N1 benchmark, NavCMPO achieves an average success rate of 74.7\%, exceeding the retrained NavDP baseline by 6.4 percentage points, while reducing inference latency from 85\,ms to 60\,ms. Experiments on a Unitree Go2 further demonstrate effective sim-to-real transfer.

cs.RO↗

arXiv2024

Does ESG Consistently Promote the Corporate Financial Performance? A Study of the Global Cruise Industry

The analysis of determinants of a company's financial performance has aroused significant attention, particularly, the environmental, social, and governance (ESG) has been the research focus in recent years. In addition to increasing revenue, the cruise industry has actively embraced the initiative of "green shipping". This study investigates the relationship between ESG and corporate financial performance (CFP) in the global cruise sector. This paper utilizes the sample data from the world's largest cruise companies over 2012-2023, to examine the ESG-CFP relationship by a regression model. The results indicate that ESG practices in cruise companies negatively influence CFP, which is further impacted by financial constraints. Furthermore, the heterogeneity analysis suggests that the high time interest earned (TIE) ratios and low total annual greenhouse gas (GHG) emissions worsen the adverse impacts of ESG on CFP. These findings contribute to the theoretical research on ESG and provide practical guidance for cruise industry operators and investors in their decision-making.

econ.GN↗