HyperAI초신경

홈 뉴스 최신 연구 논문 튜토리얼 데이터셋 백과사전 SOTA LLM 모델 GPU 랭킹 컨퍼런스

한국어

HyperAI초신경

3D Object Detection On Nuscenes Camera Only

평가 지표

Future Frame

NDS

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	Future Frame	NDS	Paper Title	Repository
Far3D	false	68.7	Far3D: Expanding the Horizon for Surround-view 3D Object Detection
BEVDet4D	false	56.9	BEVDet4D: Exploit Temporal Cues in Multi-camera 3D Object Detection
CAPE	false	62.8	CAPE: Camera View Position Embedding for Multi-View 3D Object Detection
BEVDepth-pure	false	60.9	BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detection
PETRv2-pure	false	59.2	PETRv2: A Unified Framework for 3D Perception from Multi-Camera Images
SA-BEV	false	62.4	SA-BEV: Generating Semantic-Aware Bird's-Eye-View Feature for Multi-view 3D Object Detection
SOLOFusion-pure	false	61.9	Time Will Tell: New Outlooks and A Baseline for Temporal Multi-View 3D Object Detection
StreamPETR-Large	false	67.6	Exploring Object-Centric Temporal Modeling for Efficient Multi-View 3D Object Detection
HoP	yes	68.5	Temporal Enhanced Training of Multi-view 3D Object Detector via Historical Object Prediction
BEVStereo	false	61.0	BEVStereo: Enhancing Depth Estimation in Multi-view 3D Object Detection with Dynamic Temporal Stereo
RayDN	false	68.6	Ray Denoising: Depth-aware Hard Negative Sampling for Multi-view 3D Object Detection
BEVDistill	false	59.4	BEVDistill: Cross-Modal BEV Distillation for Multi-View 3D Object Detection
PolarFormer	false	57.2	PolarFormer: Multi-camera 3D Object Detection with Polar Transformer
SparseBEV (V2-99)	yes	67.5	SparseBEV: High-Performance Sparse 3D Object Detection from Multi-Camera Videos
GeoBEV (V2-99)	false	66.2	GeoBEV: Learning Geometric BEV Representation for Multi-view 3D Object Detection
BEVFormer v2 (InternImage-XL)	yes	63.4	BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision
VCD-A	false	67.2	Leveraging Vision-Centric Multi-Modal Expertise for 3D Object Detection
SeaBird	false	59.7	SeaBird: Segmentation in Bird's View with Dice Loss Improves Monocular 3D Detection of Large Objects
BEVFormer	false	56.9	BEVFormer: Learning Bird's-Eye-View Representation from Multi-Camera Images via Spatiotemporal Transformers

0 of 19 row(s) selected.