Real2Sim2Real：用于后节眼科手术深度估计的 RetinalDepth-64K
Real2Sim2Real: RetinalDepth-64K for Depth Estimation in Posterior Segment Ophthalmic Surgery
对做单目/立体/视频深度估计的人，这是一个可复用的合成数据生成范式：用 Blender 建模解剖结构并加入折射介质与后处理，再配合时序深度方差指标评估帧间稳定性，可直接迁移到其他缺乏深度真值的显微或内窥场景。
原文依据 we introduce temporal depth variance, a novel metric for quantifying the stability of frame-to-frame depth estimation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论准确的深度估计对于后节眼科手术中的三维重建与精确导航至关重要。然而，由于深度传感器在手术显微镜下难以实际部署，获取带标注的数据仍然十分困难。为克服这一限制，我们提出 RetinalDepth，这是一个通过 Real2Sim2Real 流程构建的新型合成数据集，包含覆盖 896 个多样场景的 44,800 对时序立体图像，面向后节手术。在 Real-to-Sim 阶段，我们在 Blender 中对解剖结构准确的眼球与手术器械进行建模，融入超广角视网膜纹理、折射性房水与动态轨迹，并通过后处理增强真实感。RetinalDepth 提供同步的左右 RGB 帧，以及像素级精确的深度图、表面法线、器械分割掩码与相机参数，可支撑单目、立体与视频深度估计模型的稳健训练。在 Sim-to-Real 阶段，我们引入时序深度方差，这是一种用于量化逐帧深度估计稳定性的新指标。在 RetinalDepth 上微调可显著提升模型在合成与真实手术视频上的性能，增强泛化能力、手术精度与新手训练效果。作为首个面向后节手术的合成基准，RetinalDepth 弥合了眼科计算机视觉中的仿真到现实差距。
英文原文摘要 1406 字符
Accurate depth estimation is crucial for 3D reconstruction and precise navigation in posterior segment ophthalmic surgery. However, acquiring annotated data remains challenging due to the impracticality of depth sensors under surgical microscopes. To overcome this limitation, we introduce RetinalDepth, a novel synthetic dataset comprising 44,800 temporal stereo image pairs across 896 diverse scenes for posterior segment surgery, developed via a Real2Sim2Real pipeline. In the Real-to-Sim phase, we model anatomically accurate eyes and instruments in Blender, integrating ultra-wide-field retinal textures, refractive aqueous humor, and dynamic trajectories, enhanced by post-processing for realism. RetinalDepth provides synchronized left and right RGB frames with pixel-perfect depth maps, surface normals, instrument segmentation masks, and camera parameters, enabling robust training of monocular, stereo, and video depth estimation models. In the Sim-to-Real phase, we introduce temporal depth variance, a novel metric for quantifying the stability of frame-to-frame depth estimation. Fine-tuning on RetinalDepth significantly boosts model performance on both synthetic and real surgical videos, enhancing generalization, surgical precision, and novice training. As the first synthetic benchmark for posterior segment surgery, RetinalDepth bridges the sim-to-real gap in ophthalmic computer vision.