List your company

3D Semantic Occupancy Prediction on EmbodiedScan

18.41mIoUOcc-VLM
6.87649.870712.86515.8593Jun 18, 2026

Evaluation results

MethodLinks
Occ-VLMInput=RGB2026.0618.4149.0351.132.3135.0124.6125.8632.6737.9328.5120.9239.53
Mink-NetVenue=CVPR’19, Input=PC2026.0615.5669.9260.5251.7449.4423.0824.3345.7743.5229.7423.0239.04
SliceOccVenue=ICRA’25, Input=RGB2026.0615.4549.3251.3933.2931.6522.3925.1429.5732.7826.8319.8838.88
EmbodiedScanVenue=CVPR’24, Input=RGB2026.0614.5248.8149.5732.3232.1220.9524.327.9530.5127.319.0335.4
Cylinder3DVenue=CVPR’21, Input=PC2026.0611.5270.2263.5344.1841.5420.6317.3534.0235.6726.4815.9842.89
SurroundOccVenue=ICCV’23, Input=RGB2026.069.138.5446.1723.5523.0413.619.1527.7922.2813.1113.7224.32
OccNetVenue=ICCV’23, Input=RGB2026.068.0737.1546.925.6320.9413.1718.426.8122.8613.5913.4926.75
SPVCNNVenue=ECCV’20, Input=PC2026.067.3263.0461.338.8225.115.287.5526.616.2318.198.2625.64