Memory-efficient full-volume inference for large-scale 3D dense prediction without performance degradation

Large-volume 3D dense prediction is essential in industrial applications like energy exploration and medical image segmentation. However, existing deep learning models struggle to process full-size volumetric inputs at inference due to memory constraints and inefficient operator execution. Conventional solutions-such as tiling or compression-often introduce artifacts, compromise spatial consistency, or require retraining. Here we present a retraining-free inference optimization framework that enables accurate, efficient, whole-volume prediction without performance degradation. Our approach integrates operator spatial tiling, operator fusion, normalization statistic aggregation, and on-demand feature recomputation to reduce memory usage and accelerate runtime. Validated across multiple seismic exploration models, our framework supports full size inference on volumes exceeding 1024³ voxels. On FaultSeg3D, for instance, it completes inference on a 1024³ volume in 7.5 seconds using just 27.6 GB of memory-compared to conventional inference, which can handle only 448³ inputs under the same budget, marking a 13 × increase in volume size without loss in performance. Unlike traditional patch-wise inference, our method preserves global structural coherence, making it particularly suited for tasks inherently incompatible with chunked processing, such as implicit geological structure estimation. This work offers a generalizable, engineering-friendly solution for deploying 3D models at scale across industrial domains.