HUNTER: Unsupervised Human-centric 3D Detection via Transferring Knowledge from Synthetic Instances to Real Scenes
This work addresses the challenge of generalizing human-centric 3D detection to complex real-world scenarios for robotics applications, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackles the problem of human-centric 3D scene understanding in diverse real-life scenarios with limited labeled data by proposing an unsupervised 3D detection method that transfers knowledge from synthetic human instances to real scenes, achieving an 87.8% improvement in mAP and closely approaching fully supervised performance (62.15 mAP vs. 69.02 mAP) on the HuCenLife Dataset.
Human-centric 3D scene understanding has recently drawn increasing attention, driven by its critical impact on robotics. However, human-centric real-life scenarios are extremely diverse and complicated, and humans have intricate motions and interactions. With limited labeled data, supervised methods are difficult to generalize to general scenarios, hindering real-life applications. Mimicking human intelligence, we propose an unsupervised 3D detection method for human-centric scenarios by transferring the knowledge from synthetic human instances to real scenes. To bridge the gap between the distinct data representations and feature distributions of synthetic models and real point clouds, we introduce novel modules for effective instance-to-scene representation transfer and synthetic-to-real feature alignment. Remarkably, our method exhibits superior performance compared to current state-of-the-art techniques, achieving 87.8% improvement in mAP and closely approaching the performance of fully supervised methods (62.15 mAP vs. 69.02 mAP) on HuCenLife Dataset.