![[Cannon Selection] Generate multi-view optimal solutions in one image and combine with lora training](https://rh-images.xiaoyaoyou.com/50bd0b60897dd641e375577cd2e3a05d/2025-04-21/809c28d4a239436953708d81936ad0e3.jpg?imageMogr2/format/webp/quality/75/ignore-error/1/minisize/1)
![[Cannon Selection] Generate multi-view optimal solutions in one image and combine with lora training](https://rh-images.xiaoyaoyou.com/50bd0b60897dd641e375577cd2e3a05d/2025-04-21/58540b4ec32c4cc0bee79913354e5282.jpg?imageMogr2/format/webp/quality/75/ignore-error/1/minisize/1)
Powered By RTX 4090
One image can produce many perspectives, suitable for training lora or 3D modeling
炮老师的小课堂
2025-04-21 Update
炮老师的小课堂
2025-04-21 Update
Workflow introduction
One image can produce many perspectives, suitable for training lora or 3D modeling
Nodes Information
24
CLIPLoader
CLIPTextEncode
CLIPVisionEncode
CLIPVisionLoader
KSampler
LoadImage
LoraLoaderModelOnly
PreviewImage
UNETLoader
VAEDecodeTiled
VAELoader
GetImagesFromBatchIndexed
ImageRemoveBackground+
ImageResizeKJ
LayerMask: SegmentAnythingUltra V2
MaskPreview+
MatAnyoneVideoMatting
Note
Note Plus (mtb)
RIFE VFI
TransparentBGSession+
VHS_VideoCombine
WanImageToVideo
WanVideoTeaCacheKJ