# TiG-BEV: Multi-view BEV 3D Object Detection via Target Inner-Geometry Learning

Source: [TiG-BEV: Multi-view BEV 3D Object Detection via Target Inner-Geometry Learning](<https://www.sota2.com/research/paper/68fa6e491e209131cea3f02f>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-27T06:19:28.634Z

Authors: Peixiang Huang, Li Liu, Renrui Zhang, Song Zhang, Xinli Xu, Baichao Wang, Guoyi Liu

Published: 2022-12-28T17:53:43.000Z

## Abstract

To achieve accurate and low-cost 3D object detection, existing methods propose to benefit camera-based multi-view detectors with spatial cues provided by the LiDAR modality, e.g., dense depth supervision and bird-eye-view (BEV) feature distillation. However, they directly conduct point-to-point mimicking from LiDAR to camera, which neglects the inner-geometry of foreground targets and suffers from the modal gap between 2D-3D features. In this paper, we propose the learning scheme of Target Inner-Geometry from the LiDAR modality into camera-based BEV detectors for both dense depth and BEV features, termed as TiG-BEV. First, we introduce an inner-depth supervision module to learn the low-level relative depth relations between different foreground pixels. This enables the camera-based detector to better understand the object-wise spatial structures. Second, we design an inner-feature BEV distillation module to imitate the high-level semantics of different keypoints within foreground targets. To further alleviate the BEV feature gap between two modalities, we adopt both inter-channel and inter-keypoint distillation for feature-similarity modeling. With our target inner-geometry distillation, TiG-BEV can effectively boost BEVDepth by +2.3% NDS and +2.4% mAP, along with BEVDet by +9.1% NDS and +10.3% mAP on nuScenes val set. Code will be available at https://github.com/ADLab3Ds/TiG-BEV.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [3D Object Detection on nuScenes v1.0 (val)](<https://www.sota2.com/research/sota/3d-object-detection-on-nuscenes-v1-0-val.md>): 3D Object Detection; nuScenes v1.0 (val); mAP (Overall): 33.8.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2212.13979>)

[PDF](<https://arxiv.org/pdf/2212.13979.pdf>)