NeRF-Det++: Incorporating Semantic Cues and Perspective-aware Depth Supervision for Indoor Multi-View 3D Detection

TL;DR

NeRF-Det++ enhances indoor multi-view 3D detection by integrating semantic cues, perspective-aware sampling, and ordinal depth supervision, leading to improved accuracy over previous methods.

Contribution

It introduces three novel solutions—semantic enhancement, perspective-aware sampling, and ordinal residual depth supervision—to address key shortcomings in existing NeRF-Det based 3D detection methods.

Findings

Outperforms NeRF-Det by +1.9% mAP@0.25 on ScanNetV2

Achieves +3.5% mAP@0.50 on ScanNetV2

Demonstrates effectiveness on ScanNetV2 and ARKITScenes datasets

Abstract

NeRF-Det has achieved impressive performance in indoor multi-view 3D detection by innovatively utilizing NeRF to enhance representation learning. Despite its notable performance, we uncover three decisive shortcomings in its current design, including semantic ambiguity, inappropriate sampling, and insufficient utilization of depth supervision. To combat the aforementioned problems, we present three corresponding solutions: 1) Semantic Enhancement. We project the freely available 3D segmentation annotations onto the 2D plane and leverage the corresponding 2D semantic maps as the supervision signal, significantly enhancing the semantic awareness of multi-view detectors. 2) Perspective-aware Sampling. Instead of employing the uniform sampling strategy, we put forward the perspective-aware sampling policy that samples densely near the camera while sparsely in the distance, more effectively collecting the valuable geometric clues. 3)Ordinal Residual Depth Supervision. As opposed to directly regressing the depth values that are difficult to optimize, we divide the depth range of each scene into a fixed number of ordinal bins and reformulate the depth prediction as the combination of the classification of depth bins as well as the regression of the residual depth values, thereby benefiting the depth learning process. The resulting algorithm, NeRF-Det++, has exhibited appealing performance in the ScanNetV2 and ARKITScenes datasets. Notably, in ScanNetV2, NeRF-Det++ outperforms the competitive NeRF-Det by +1.9% in mAP@0.25 and +3.5% in mAP@0.50$. The code will be publicly at https://github.com/mrsempress/NeRF-Detplusplus.