PointTransformerV3

class PointTransformerV3(in_channels=6, order=('z', 'z-trans'), stride=(2, 2, 2, 2), enc_depths=(2, 2, 2, 6, 2), enc_channels=(32, 64, 128, 256, 512), enc_num_head=(2, 4, 8, 16, 32), enc_patch_size=(48, 48, 48, 48, 48), dec_depths=(2, 2, 2, 2), dec_channels=(64, 64, 128, 256), dec_num_head=(4, 4, 8, 16), dec_patch_size=(48, 48, 48, 48), mlp_ratio=4, qk_norm=False, qkv_bias=True, qk_scale=None, attn_drop=0.0, proj_drop=0.0, drop_path=0.3, layer_scale=None, pre_norm=True, shuffle_orders=True, enable_rpe=False, enable_flash=True, enable_cpe=True, upcast_attention=False, upcast_softmax=False, traceable=False, mask_token=False, enc_mode=False, freeze_encoder=False, cpe_first_layer_only=False, enc_cpe_channels=(96, 96, 96, 96, 96), dec_cpe_channels=(96, 96, 96, 96), embedding_act_layer=True, add_iso_features=False, iso_k=8, iso_r=None, rope_base=10, rope_jitter=None, rope_rescale=None)

Bases: PointModule

Point Transformer V3 with 3D RoPE and bottleneck CPE.

PointTransformerV3.forward(data_dict)