RoPETR claims 70.9% NDS on nuScenes by applying M-RoPE rotary embeddings to BEV coordinates and temporal IDs in StreamPETR, but a missing RayDN ablation leaves the attribution unclear.
Ray denoising: Depth-aware hard negative sampling for multi-view 3d object detection
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
RoPETR: Improving Temporal Camera-Only 3D Detection by Integrating Enhanced Rotary Position Embedding
RoPETR claims 70.9% NDS on nuScenes by applying M-RoPE rotary embeddings to BEV coordinates and temporal IDs in StreamPETR, but a missing RayDN ablation leaves the attribution unclear.