TransNormal-2：以幾何為基礎的矯正流與邊緣感知解碼用於精確法線估計

摘要

基於擴散模型的方法可實現單目幾何估計，然而其像素空間精度受到一個共同且研究不足的誤差來源所限制：VAE 重建退化。VAE 編碼器-解碼器中的 8 倍空間壓縮會使物件邊界處的表面法線退化；即使僅將真值法線編碼再解碼，也會引入 1.3–8.5° 的平均角度誤差（MAE），而邊緣 MAE 更達全域 MAE 的 2.8 倍。我們提出 TransNormal-2，一個基於 FLUX.2 的修正流框架，具備單步確定性推論，能在 VAE 解碼器的前後兩端處理此退化：一是在訓練期間如何監督潛在預測，二是在推論時如何修正解碼後的法線。首先，幾何感知的像素空間損失，包括逆渲染自一致性、von Mises-Fisher 角度損失，以及小波邊緣感知正則化，透過在 VAE 解碼後強制球面法線幾何與漫反射成像線索，補足潛在 MSE。其次，輕量級幾何精煉模組（GRM）施加 RGB 引導的殘差修正，以降低邊界局部化的解碼誤差，同時不任意重寫粗略預測。在一般場景基準上，TransNormal-2 在所有八項報告指標上達到或超越 MoGe-2，同時僅使用相當於其 1.4% 的任務特定法線標註。對透明物體的增益最為明顯，相較於最強先前基線，在 ClearGrasp 上將 MAE 降低 4.2°，在 ClearPose 上降低 3.1°。程式碼將於 https://longxiang-ai.github.io/TransNormal-2 發布。

English

Diffusion-based models enable monocular geometry estimation, yet their pixel-space precision is limited by a shared, under-studied error source: VAE reconstruction degradation. The 8x spatial compression in the VAE encoder-decoder degrades surface normals at object boundaries; even encoding and decoding ground-truth normals introduces 1.3--8.5° of mean angular error (MAE), with edge MAE reaching 2.8x the global MAE. We present TransNormal-2, a FLUX.2-based rectified-flow framework with single-step deterministic inference that addresses this degradation on both sides of the VAE decoder: in how latent predictions are supervised during training, and in how decoded normals are corrected at inference. First, geometry-aware pixel-space losses, including inverse rendering self-consistency, von~Mises-Fisher angular loss, and wavelet edge-aware regularization, complement latent MSE by enforcing spherical normal geometry and diffuse image-formation cues after VAE decoding. Second, a lightweight Geometric Refinement Module (GRM) applies an RGB-guided residual correction to reduce boundary-localized decoding errors without freely rewriting the coarse prediction. On general-scene benchmarks, TransNormal-2 matches or exceeds MoGe-2 on all eight reported metrics while using only 1.4% as many task-specific normal annotations. The gains are clearest for transparent objects, reducing MAE by 4.2° on ClearGrasp and 3.1° on ClearPose over the strongest prior baselines. Code will be released at https://longxiang-ai.github.io/TransNormal-2.