The paper presents a pipeline for constructing accurate ground‑truth annotations for visual tracking across the air–water interface, synchronizing camera frames with motion‑capture data and correcting underwater projections. It introduces a Cross‑Medium Recovery Policy (CMRP) that uses confidence‑triggered template selection to improve MixFormerV2’s tracking performance without retraining the backbone. CMRP raises Macro Success AUC by 2.95 points and increases tracking coverage while reducing recovery latency.
Visual tracking across the air-water interface is challenged by splashes, bubbles, refraction, reflections, and abrupt appearance changes that can temporarily invalidate observations. This setting pos...
arXiv:2609.22353v1 Announce Type: new Abstract: Mobile river monitoring robots must interpret obstacles and water boundaries that geographic waypoints alone cannot describe. On resource constrained p...
RiverVLN introduces the first benchmark for long‑horizon vision‑language navigation (VLN) of unmanned surface vehicles (USVs) in continuous riverine motion. The PGT‑NAV framework converts navigation instructions into an ordered sequence of visually verifiable semantic phases, maintaining an active phase online through grounded visual and motion evidence. This phase‑grounded approach reduces recursive position and heading drift, achieving a 0.79 success rate in Unity‑ROS closed‑loop tests and demonstrating transfer to real‑world USV deployment.
arXiv:2608. 19710v1 Announce Type: cross Abstract: Reliable underwater robotic perception remains difficult because optical imagery degrades under turbidity, wavelength-dependent attenuation, low illumination, scattering, and blur.
Underwater C³-JEPA is an object‑centric, cross‑view predictive world model designed for near‑field heavy‑load ROV salvage. It encodes synchronized multi‑camera RGB observations and vehicle control signals into task‑object and context tokens, fuses cross‑camera evidence via held‑out‑view attention, and predicts future states conditioned on control without using contact sensors. The model demonstrates superior transfer of task‑relevant information to downstream probes compared to a reconstruction‑free baseline, supports model‑predictive control and imagined‑rollout training, and validates its effectiveness on real underwater video by accurately recovering withheld camera states and maintaining predictive lead over persistence.
AquaBEV is a monocular underwater occupancy model that predicts local bird’s‑eye‑view (BEV) occupancy from a single RGB image. It uses paired 3D imaging sonar data as geometric supervision during training, mapping visual features into a calibration‑free polar representation and decoding along the range dimension before reconstructing Cartesian BEV coordinates. In a controlled underwater occupancy benchmark, AquaBEV outperforms the strongest transferred baseline with 31.4 % Visible IoU and 38.6 % Observed IoU, achieving 4.0 % and 4.3 % relative improvements respectively.
The paper investigates the performance of Gaussian splatting techniques for 3D reconstruction in underwater settings, evaluating five publicly available systems across datasets with varying turbidity, illumination loss, and colour attenuation, as well as an industrial survey. The study finds that reconstruction quality is more influenced by environmental setup—such as water clarity and illumination geometry—than by the specific algorithmic architecture, with clear water yielding high frame registration rates and artificial moving lights favoring medium‑blind splatting. The authors release all scene builds, per‑run configurations, and evaluation code to support reproducibility.
The underwater environment is challenging for 3D reconstruction, because particles suspended in the water scatter and diffuse light, turbidity varies, absorption depends on wavelength, and illuminatio...
Reconstructing photorealistic scenes in unconstrained underwater environments remains challenging due to severe media-induced light scattering and unpredictable dynamic objects. Recent feed-forward vi...
arXiv:2608.21402v1 Announce Type: cross Abstract: World action models (WAMs) jointly denoise future video frames and robot actions, and the video prior is expected to generalize their control. Camera...
arXiv:2608.22906v1 Announce Type: new Abstract: Recent monocular 3D Gaussian Splatting (3DGS) streaming reconstruction methods have achieved impressive performance by balancing reconstruction quality...
arXiv:2608. 06154v1 Announce Type: cross Abstract: Vision-language models (VLMs) are increasingly used as zero-shot controllers, but successful trajectories do not necessarily show that decisions are grounded in visual input: simulator dynamics and conservative action priors can produce favourable scores without meaningful perception.