arXiv subjects

Yihan Fang

Publications and source records attributed to Yihan Fang.

4 recordsLinked to original sources

arXiv2026

PointPilot: Unified Robotic Manipulation via Object-Centric 3D Point Flow Prediction

Generalization remains a central challenge in robotics, calling for policies that address domain variations from multiple dimensions such as diverse embodiments, camera poses, and/or scene layouts. In order to approach this problem, we introduce PointPilot, a task-focused and 3D aligned manipulation framework built on a 3D space point-flow prediction. Rather than modeling the global scene, PointPilot only represents task-focused objects (robot grippers treated as objects), avoiding the prediction of potentially redundant scene parts unrelated to the manipulation tasks. Moreover, PointPilot aligns 3D spatial representation in one unified coordinate and supports closed-loop and fully automatic action execution across different manipulation modes from object point flow without a separately trained action expert, while existing works typically need an extra action head or are (semi-)automatic and single-arm mode specific. Given sensor observations and language instructions, our model jointly learns subtask planning, task-relevant object grounding, and 3D point flow prediction in an end-to-end way. By processing heterogeneous data in a shared 3D point flow space, PointPilot is able to absorb demonstrations from different embodiments. Extensive simulative and real world experiments validate our method. Notably, we demonstrate zero-shot embodiment generalization with PointPilot. A single trained policy can be deployed on previously unseen embodiments and successfully complete manipulation tasks without any additional fine-tuning.

cs.RO↗

arXiv2026

UNITAS: A 3D-Native World Action Model for Embodied Manipulation

World action models (WAMs) aim to answer a coupled physical question: given a task instruction, what motion should the robot execute, and how will that motion change the surrounding world? Most existing WAMs build on pretrained video generators and represent world evolution through images or visual latents. Robotic interaction, however, takes place in metric three-dimensional space, while images are view-dependent projections whose pixel distances do not directly encode physical distances. We introduce UNITAS, to our knowledge the first 3D-native world action model that unifies observations, actions, and scene dynamics in a shared metric 3D frame within each interaction, using a common representation across robot embodiments and human hands. Action flow represents human hands and robot grippers as 3D point trajectories, while scene flow describes scene-point displacements conditioned on these trajectories. World-aligned 3D positional embeddings ground visual tokens with or without depth input, and a physical-time trajectory tokenizer encodes each point trajectory as one token anchored at its current 3D position. This interface supports both direct action execution and action-conditioned scene prediction. With 1.7B parameters, UNITAS achieves the best action-conditioned scene prediction on RoboTwin among the compared methods, with up to 49% lower displacement errors than PointWorld, and state-of-the-art manipulation success, including 99.8% on LIBERO and an average of 85% across real-world tasks. The code is available at https://github.com/DexForce/UNITAS.

cs.RO↗

arXiv2026

Y-Trim: Evidence-gated Adaptase tail trimming for single-stranded bisulfite sequencing

Background: Single-stranded whole-genome bisulfite sequencing (ssWGBS) enables DNA methylation profiling in low-input and highly fragmented material, including cell-free DNA. In widely used post-bisulfite protocols, Adaptase-mediated tailing adds stochastic, template-free end sequence. Unlike adapter-defined junctions, these tails lack a fixed sequence template, so trimming must be decided from FASTQ-stage observables under intrinsic uncertainty. Results: We show that bisulfite-induced compositional degeneracy implies a strictly positive error floor for any fixed per-read boundary rule under a finite nucleotide alphabet. Guided by this limit, we introduce Y-Trim, an evidence-gated framework that separates admission (should we trim) from inference (where to trim). For Read 2, Y-Trim performs per-read adaptive cut placement via a fixed, chemistry-typed matrix-linear texture scoring scheme; for Read 1, it uses automated sample-level anchoring when read-level localization is feasibility-limited. Across modules, Y-Trim is an explicit, chemistry-specific decision rule with interpretable operating points. On a curated 34-run public cohort (CCGB-34) and simulator stress tests with known latent boundaries, Y-Trim exhibits stable Read 2 operating behavior and Read 1 feasibility-limited behavior consistent with conditional read-through. Conclusions: Template-free Adaptase tail trimming is best viewed as an evidence-limited FASTQ-stage decision rather than a generic preprocessing knob. By making admissibility and abstention explicit and exposing interpretable genomic-retention versus residual-carryover trade-offs, Y-Trim provides a practical uncertainty-aware preprocessing strategy for ssWGBS.

q-bio.GN↗

arXiv2025

BYOCL: Build Your Own Consistent Latent with Hierarchical Representative Latent Clustering

To address the semantic inconsistency issue with SAM or other single-image segmentation models handling image sequences, we introduce BYOCL. This novel model outperforms SAM in extensive experiments, showcasing its Hierarchical prototype capabilities across CLIP and other representations. BYOCL significantly reduces time and space consumption by dividing inputs into smaller batches, achieving exponential time reduction compared to previous methods. Our approach leverages the SAM image encoder for feature extraction, followed by Intra-Batch and Inter-Batch clustering algorithms. Extensive experiments demonstrate that BYOCL far exceeds the previous state-of-the-art single image segmentation model. Our work is the first to apply consistent segmentation using foundation models without requiring training, utilizing plug-and-play modules for any latent space, making our method highly efficientModels are available at \href{https://github.com/cyt1202/BYOCL.git

cs.CV↗