Open-vocabulary 3D object detection with promptable segmentation

Author1, Author2, Author3, Author4, Author5

3D object detectionpromptable segmentationautonomous drivingopen-vocabulary

2609.19358

Builder Relevance

80%

2h ago

Abstract

This paper explores training-free, open-vocabulary 3D object detection using a promptable segmentation model to generate instance masks from text prompts.

Reality Card

Core Claim

The proposed method achieves a mean average precision (mAP) of 0.413 and a nuScenes detection score (NDS) of 0.555 by utilizing promptable segmentation for 3D object detection without the need for extensive training.

Method / Result

The method reaches 0.413 mAP / 0.555 NDS with zero labeling cost.

Limitations

The main limitation is the reliance on the accuracy of class naming and geometric precision, which affects measurement outcomes.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.

← Back to all papers