"""Hand-specific, side-conditioned VAE; 15 local rotations, no wrist/global/shape.""" import torch from torch import nn from torch.nn import functional as F def aa_to_matrix(x): theta=x.norm(dim=-1,keepdim=True);x0,x1,x2=x.unbind(-1);z=torch.zeros_like(x0) k=torch.stack([z,-x2,x1,x2,z,-x0,-x1,x0,z],-1).reshape(*x.shape[:-1],3,3) a=torch.sinc(theta/torch.pi)[...,None];b=(.5*torch.sinc(theta/(2*torch.pi)).square())[...,None] return torch.eye(3,device=x.device,dtype=x.dtype)+a*k+b*(k@k) def matrix_to_aa(m): m00,m01,m02=m[...,0,0],m[...,0,1],m[...,0,2];m10,m11,m12=m[...,1,0],m[...,1,1],m[...,1,2];m20,m21,m22=m[...,2,0],m[...,2,1],m[...,2,2] qabs=torch.stack([1+m00+m11+m22,1+m00-m11-m22,1-m00+m11-m22,1-m00-m11+m22],-1).clamp_min(1e-8).sqrt() candidates=torch.stack([torch.stack([qabs[...,0]**2,m21-m12,m02-m20,m10-m01],-1),torch.stack([m21-m12,qabs[...,1]**2,m10+m01,m02+m20],-1),torch.stack([m02-m20,m10+m01,qabs[...,2]**2,m12+m21],-1),torch.stack([m10-m01,m20+m02,m21+m12,qabs[...,3]**2],-1)],-2)/(2*qabs[...,None].clamp_min(.1)) ix=qabs.argmax(-1);q=candidates.gather(-2,ix[...,None,None].expand(*ix.shape,1,4)).squeeze(-2);q=F.normalize(q,dim=-1);q=torch.where(q[...,:1]<0,-q,q) norm=q[...,1:].norm(dim=-1,keepdim=True);angle=2*torch.atan2(norm,q[...,:1]);scale=torch.where(norm>1e-6,angle/norm.clamp_min(1e-8),2+norm.square()/3) return q[...,1:]*scale def matrix_to_six(m):return m[...,:2,:].reshape(*m.shape[:-2],6) def six_to_matrix(x): a,b=x[...,:3],x[...,3:];r0=F.normalize(a,dim=-1,eps=1e-7);r1=F.normalize(b-(r0*b).sum(-1,keepdim=True)*r0,dim=-1,eps=1e-7);r2=torch.cross(r0,r1,dim=-1);return torch.stack([r0,r1,r2],-2) class HandVAE(nn.Module): def __init__(self,latent_dim=20,width=256): super().__init__();self.latent_dim=latent_dim;self.width=width self.register_buffer('input_mean',torch.zeros(90));self.register_buffer('input_std',torch.ones(90)) self.encoder=nn.Sequential(nn.Linear(92,width),nn.LeakyReLU(.2),nn.Linear(width,width),nn.LeakyReLU(.2)) self.mu=nn.Linear(width,latent_dim);self.logvar=nn.Linear(width,latent_dim);nn.init.constant_(self.logvar.bias,-2) self.decoder=nn.Sequential(nn.Linear(latent_dim+2,width),nn.LeakyReLU(.2),nn.Linear(width,width),nn.LeakyReLU(.2),nn.Linear(width,90)) def encode(self,rot,side): x=(matrix_to_six(rot).flatten(-2)-self.input_mean)/self.input_std;h=self.encoder(torch.cat([x,F.one_hot(side.long(),2).to(x)],-1));return self.mu(h),self.logvar(h).clamp(-8,4) def decode(self,z,side): x=self.decoder(torch.cat([z,F.one_hot(side.long(),2).to(z)],-1))*self.input_std+self.input_mean;return six_to_matrix(x.reshape(*z.shape[:-1],15,6)) def forward(self,rot,side): mu,lv=self.encode(rot,side);z=mu+torch.randn_like(mu)*(.5*lv).exp() if self.training else mu;return self.decode(z,side),mu,lv def load_vae(path,device='cuda'): ck=torch.load(path,map_location=device,weights_only=False);m=HandVAE(**ck['architecture']).to(device);m.load_state_dict(ck['state_dict'],strict=True);m.eval();m.requires_grad_(False);return m,ck