3D Object Detection On Scannetv2

Metriken

mAP@0.25

mAP@0.5

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Vergleichstabelle

Modellname	mAP@0.25	mAP@0.5
Modell 1	74.6	65.3
tr3d-towards-real-time-indoor-3d-object	72.9	59.3
rbgnet-ray-based-grouping-for-3d-object	70.6	55.2
a-hierarchical-graph-network-for-3d-object	61.3	34.4
divide-and-conquer-3d-point-cloud-instance	69.3	60.1
v-detr-detr-with-vertex-relative-position	77.8	65.9
point-gcc-universal-self-supervised-3d-scene	73.1	59.6
back-tracing-representative-points-for-voting	66.1	50.9
softgroup-for-3d-instance-segmentation-on	71.6	59.4
generative-sparse-detection-networks-for-3d	62.8	34.8
spgroup3d-superpoint-grouping-network-for	74.3	59.6
h3dnet-3d-object-detection-using-hybrid	67.2	48.1
imgeonet-image-induced-geometry-aware-voxel	54.8	28.4
deep-hough-voting-for-3d-object-detection-in	58.6	33.5
group-free-3d-object-detection-via	69.1	52.8
swin3d-a-pretrained-transformer-backbone-for	76.4	63.2
oneformer3d-one-transformer-for-unified-point	76.9	65.3
multimodal-token-fusion-for-vision	70.8	54.2
unidet3d-multi-dataset-indoor-3d-object	77.5	66.1
state-space-model-meets-transformer-a-new-1	78.8	67.9
sgpn-similarity-group-proposal-network-for-3d	20.7	-
imvoxelnet-image-to-voxels-projection-for	48.1	22.7
3d-mpa-multi-proposal-aggregation-for-3d	64.2	49.2
point-cloud-pre-training-with-natural-3d	63.4	39.9
surface-representation-for-point-clouds	71.2	54.8
an-end-to-end-transformer-model-for-3d-object	65.0	47.0
query-refinement-transformer-for-3d-instance	73.4	61.7
3d-sis-3d-semantic-instance-segmentation-of	40.2	22.5
fcaf3d-fully-convolutional-anchor-free-3d	71.5	57.3
cagroup3d-class-aware-grouping-for-3d-object	75.1	61.3
mv-detr-multi-modality-indoor-object	78.0	65.8
gspn-generative-shape-proposal-network-for-3d	30.6	17.7
state-space-model-meets-transformer-a-new-1	71.3	58.1