# Distilling Future Temporal Knowledge with Masked Feature Reconstruction for 3D Object Detection

Source: [Distilling Future Temporal Knowledge with Masked Feature Reconstruction for 3D Object Detection](<https://www.sota2.com/research/paper/6938febb6cfd488761a53ecc>)

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

Data updated: 2025-12-11T05:03:10.792Z

Authors: Haowen Zheng, Hu Zhu, Lu Deng, Weihao Gu, Yang Yang, Yanyan Liang

Published: 2025-12-09T05:01:29.000Z

## Abstract

Camera-based temporal 3D object detection has shown impressive results in autonomous driving, with offline models improving accuracy by using future frames. Knowledge distillation (KD) can be an appealing framework for transferring rich information from offline models to online models. However, existing KD methods overlook future frames, as they mainly focus on spatial feature distillation under strict frame alignment or on temporal relational distillation, thereby making it challenging for online models to effectively learn future knowledge. To this end, we propose a sparse query-based approach, Future Temporal Knowledge Distillation (FTKD), which effectively transfers future frame knowledge from an offline teacher model to an online student model. Specifically, we present a future-aware feature reconstruction strategy to encourage the student model to capture future features without strict frame alignment. In addition, we further introduce future-guided logit distillation to leverage the teacher's stable foreground and background context. FTKD is applied to two high-performing 3D object detection baselines, achieving up to 1.3 mAP and 1.3 NDS gains on the nuScenes dataset, as well as the most accurate velocity estimation, without increasing inference cost.

## 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-trainval (val)](<https://www.sota2.com/research/sota/3d-object-detection-on-nuscenes-v1-0-trainval-val.md>): 3D Object Detection; nuScenes v1.0-trainval (val); NDS: 56.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2512.08247>)

[PDF](<https://arxiv.org/pdf/2512.08247.pdf>)