Yangtao Hu | Searcharxiv

arXiv2026

Orca: The World is in Your Mind

We introduce Orca, an initial instantiation of a general world foundation model. Orca learns a unified world latent space from multimodal world signals and exposes it through multimodal readout interfaces. Rather than optimizing isolated next-token, next-frame, or next-action prediction, we are centered on Next-State-Prediction modeling, offering a unified state-transition modeling route toward understanding, predicting, and acting upon the world. Orca learns through two complementary paradigms: unconscious learning captures dense natural state transitions from continuous videos, and conscious learning models sparse meaningful state transitions by language-described events and VQA supervision. For pre-training, we construct a large-scale world-learning inventory data, including 125K hours of video data and 160M event annotations. After pre-training, Orca learns a unified world latent space. To examine whether the learned latent supports downstream, we evaluate it by three representative downstream readouts: text generation, image prediction, and embodied action generation. Orca's backbone is frozen, and only the lightweight modality-specific decoders are trainable. Experiments show the scalability of the proposed paradigm and verify that stronger world latent enables stronger downstream readouts. Orca outperforms similar-sized specialized baselines. These results show that Orca, as a general world foundation model, presents a promising approach to understanding, predicting, and acting upon the world. Finally, we discuss the current limitations, aiming to provide useful insights and inspiration for the community.

cs.CV↗

arXiv2020

Channel Attention Residual U-Net for Retinal Vessel Segmentation

Retinal vessel segmentation is a vital step for the diagnosis of many early eye-related diseases. In this work, we propose a new deep learning model, namely Channel Attention Residual U-Net (CAR-UNet), to accurately segment retinal vascular and non-vascular pixels. In this model, we introduced a novel Modified Efficient Channel Attention (MECA) to enhance the discriminative ability of the network by considering the interdependence between feature maps. On the one hand, we apply MECA to the "skip connections" in the traditional U-shaped networks, instead of simply copying the feature maps of the contracting path to the corresponding expansive path. On the other hand, we propose a Channel Attention Double Residual Block (CADRB), which integrates MECA into a residual structure as a core structure to construct the proposed CAR-UNet. The results show that our proposed CAR-UNet has reached the state-of-the-art performance on three publicly available retinal vessel datasets: DRIVE, CHASE DB1 and STARE.

eess.IV↗