import torch from pprint import pprint from metrics.evaluation_metrics import jsd_between_point_cloud_sets as JSD from metrics.evaluation_metrics import compute_all_metrics, EMD_CD import torch.nn as nn import torch.optim as optim import torch.utils.data import argparse from model.unet import get_model from torch.distributions import Normal from utils.file_utils import * from utils.visualize import * from utils.mitsuba_renderer import write_to_xml_batch from model.pvcnn_generation import PVCNN2Base from tqdm import tqdm from datasets.shapenet_data_pc import ShapeNet15kPointClouds ''' models ''' def normal_kl(mean1, logvar1, mean2, logvar2): """ KL divergence between normal distributions parameterized by mean and log-variance. """ return 0.5 * (-1.0 + logvar2 - logvar1 + torch.exp(logvar1 - logvar2) + (mean1 - mean2)**2 * torch.exp(-logvar2)) def discretized_gaussian_log_likelihood(x, *, means, log_scales): # Assumes data is integers [0, 1] assert x.shape == means.shape == log_scales.shape px0 = Normal(torch.zeros_like(means), torch.ones_like(log_scales)) centered_x = x - means inv_stdv = torch.exp(-log_scales) plus_in = inv_stdv * (centered_x + 0.5) cdf_plus = px0.cdf(plus_in) min_in = inv_stdv * (centered_x - .5) cdf_min = px0.cdf(min_in) log_cdf_plus = torch.log(torch.max(cdf_plus, torch.ones_like(cdf_plus)*1e-12)) log_one_minus_cdf_min = torch.log(torch.max(1. - cdf_min, torch.ones_like(cdf_min)*1e-12)) cdf_delta = cdf_plus - cdf_min log_probs = torch.where( x < 0.001, log_cdf_plus, torch.where(x > 0.999, log_one_minus_cdf_min, torch.log(torch.max(cdf_delta, torch.ones_like(cdf_delta)*1e-12)))) assert log_probs.shape == x.shape return log_probs class GaussianDiffusion: def __init__(self,betas, loss_type, model_mean_type, model_var_type): self.loss_type = loss_type self.model_mean_type = model_mean_type self.model_var_type = model_var_type assert isinstance(betas, np.ndarray) self.np_betas = betas = betas.astype(np.float64) # computations here in float64 for accuracy assert (betas > 0).all() and (betas <= 1).all() timesteps, = betas.shape self.num_timesteps = int(timesteps) # initialize twice the actual length so we can keep running for eval # betas = np.concatenate([betas, np.full_like(betas[:int(0.2*len(betas))], betas[-1])]) alphas = 1. - betas alphas_cumprod = torch.from_numpy(np.cumprod(alphas, axis=0)).float() alphas_cumprod_prev = torch.from_numpy(np.append(1., alphas_cumprod[:-1])).float() self.betas = torch.from_numpy(betas).float() self.alphas_cumprod = alphas_cumprod.float() self.alphas_cumprod_prev = alphas_cumprod_prev.float() # calculations for diffusion q(x_t | x_{t-1}) and others self.sqrt_alphas_cumprod = torch.sqrt(alphas_cumprod).float() self.sqrt_one_minus_alphas_cumprod = torch.sqrt(1. - alphas_cumprod).float() self.log_one_minus_alphas_cumprod = torch.log(1. - alphas_cumprod).float() self.sqrt_recip_alphas_cumprod = torch.sqrt(1. / alphas_cumprod).float() self.sqrt_recipm1_alphas_cumprod = torch.sqrt(1. / alphas_cumprod - 1).float() betas = torch.from_numpy(betas).float() alphas = torch.from_numpy(alphas).float() # calculations for posterior q(x_{t-1} | x_t, x_0) posterior_variance = betas * (1. - alphas_cumprod_prev) / (1. - alphas_cumprod) # above: equal to 1. / (1. / (1. - alpha_cumprod_tm1) + alpha_t / beta_t) self.posterior_variance = posterior_variance # below: log calculation clipped because the posterior variance is 0 at the beginning of the diffusion chain self.posterior_log_variance_clipped = torch.log(torch.max(posterior_variance, 1e-20 * torch.ones_like(posterior_variance))) self.posterior_mean_coef1 = betas * torch.sqrt(alphas_cumprod_prev) / (1. - alphas_cumprod) self.posterior_mean_coef2 = (1. - alphas_cumprod_prev) * torch.sqrt(alphas) / (1. - alphas_cumprod) @staticmethod def _extract(a, t, x_shape): """ Extract some coefficients at specified timesteps, then reshape to [batch_size, 1, 1, 1, 1, ...] for broadcasting purposes. """ bs, = t.shape assert x_shape[0] == bs out = torch.gather(a, 0, t) assert out.shape == torch.Size([bs]) return torch.reshape(out, [bs] + ((len(x_shape) - 1) * [1])) def q_mean_variance(self, x_start, t): mean = self._extract(self.sqrt_alphas_cumprod.to(x_start.device), t, x_start.shape) * x_start variance = self._extract(1. - self.alphas_cumprod.to(x_start.device), t, x_start.shape) log_variance = self._extract(self.log_one_minus_alphas_cumprod.to(x_start.device), t, x_start.shape) return mean, variance, log_variance def q_sample(self, x_start, t, noise=None): """ Diffuse the data (t == 0 means diffused for 1 step) """ if noise is None: noise = torch.randn(x_start.shape, device=x_start.device) assert noise.shape == x_start.shape return ( self._extract(self.sqrt_alphas_cumprod.to(x_start.device), t, x_start.shape) * x_start + self._extract(self.sqrt_one_minus_alphas_cumprod.to(x_start.device), t, x_start.shape) * noise ) def q_posterior_mean_variance(self, x_start, x_t, t): """ Compute the mean and variance of the diffusion posterior q(x_{t-1} | x_t, x_0) """ assert x_start.shape == x_t.shape posterior_mean = ( self._extract(self.posterior_mean_coef1.to(x_start.device), t, x_t.shape) * x_start + self._extract(self.posterior_mean_coef2.to(x_start.device), t, x_t.shape) * x_t ) posterior_variance = self._extract(self.posterior_variance.to(x_start.device), t, x_t.shape) posterior_log_variance_clipped = self._extract(self.posterior_log_variance_clipped.to(x_start.device), t, x_t.shape) assert (posterior_mean.shape[0] == posterior_variance.shape[0] == posterior_log_variance_clipped.shape[0] == x_start.shape[0]) return posterior_mean, posterior_variance, posterior_log_variance_clipped def p_mean_variance(self, denoise_fn, data, t, clip_denoised: bool, return_pred_xstart: bool): model_output = denoise_fn(data, t) if self.model_var_type in ['fixedsmall', 'fixedlarge']: # below: only log_variance is used in the KL computations model_variance, model_log_variance = { # for fixedlarge, we set the initial (log-)variance like so to get a better decoder log likelihood 'fixedlarge': (self.betas.to(data.device), torch.log(torch.cat([self.posterior_variance[1:2], self.betas[1:]])).to(data.device)), 'fixedsmall': (self.posterior_variance.to(data.device), self.posterior_log_variance_clipped.to(data.device)), }[self.model_var_type] model_variance = self._extract(model_variance, t, data.shape) * torch.ones_like(data) model_log_variance = self._extract(model_log_variance, t, data.shape) * torch.ones_like(data) else: raise NotImplementedError(self.model_var_type) if self.model_mean_type == 'eps': x_recon = self._predict_xstart_from_eps(data, t=t, eps=model_output) if clip_denoised: x_recon = torch.clamp(x_recon, -.5, .5) model_mean, _, _ = self.q_posterior_mean_variance(x_start=x_recon, x_t=data, t=t) else: raise NotImplementedError(self.loss_type) assert model_mean.shape == x_recon.shape == data.shape assert model_variance.shape == model_log_variance.shape == data.shape if return_pred_xstart: return model_mean, model_variance, model_log_variance, x_recon else: return model_mean, model_variance, model_log_variance def _predict_xstart_from_eps(self, x_t, t, eps): assert x_t.shape == eps.shape return ( self._extract(self.sqrt_recip_alphas_cumprod.to(x_t.device), t, x_t.shape) * x_t - self._extract(self.sqrt_recipm1_alphas_cumprod.to(x_t.device), t, x_t.shape) * eps ) ''' samples ''' def p_sample(self, denoise_fn, data, t, noise_fn, clip_denoised=False, return_pred_xstart=False, use_var=True): """ Sample from the model """ model_mean, _, model_log_variance, pred_xstart = self.p_mean_variance(denoise_fn, data=data, t=t, clip_denoised=clip_denoised, return_pred_xstart=True) noise = noise_fn(size=data.shape, dtype=data.dtype, device=data.device) assert noise.shape == data.shape # no noise when t == 0 nonzero_mask = torch.reshape(1 - (t == 0).float(), [data.shape[0]] + [1] * (len(data.shape) - 1)) sample = model_mean if use_var: sample = sample + nonzero_mask * torch.exp(0.5 * model_log_variance) * noise assert sample.shape == pred_xstart.shape return (sample, pred_xstart) if return_pred_xstart else sample def p_sample_loop(self, denoise_fn, shape, device, noise_fn=torch.randn, constrain_fn=lambda x, t:x, clip_denoised=True, max_timestep=None, keep_running=False): """ Generate samples keep_running: True if we run 2 x num_timesteps, False if we just run num_timesteps """ if max_timestep is None: final_time = self.num_timesteps else: final_time = max_timestep assert isinstance(shape, (tuple, list)) img_t = noise_fn(size=shape, dtype=torch.float, device=device) for t in reversed(range(0, final_time if not keep_running else len(self.betas))): img_t = constrain_fn(img_t, t) t_ = torch.empty(shape[0], dtype=torch.int64, device=device).fill_(t) img_t = self.p_sample(denoise_fn=denoise_fn, data=img_t,t=t_, noise_fn=noise_fn, clip_denoised=clip_denoised, return_pred_xstart=False).detach() assert img_t.shape == shape return img_t def reconstruct(self, x0, t, denoise_fn, noise_fn=torch.randn, constrain_fn=lambda x, t:x): assert t >= 1 t_vec = torch.empty(x0.shape[0], dtype=torch.int64, device=x0.device).fill_(t-1) encoding = self.q_sample(x0, t_vec) img_t = encoding for k in reversed(range(0,t)): img_t = constrain_fn(img_t, k) t_ = torch.empty(x0.shape[0], dtype=torch.int64, device=x0.device).fill_(k) img_t = self.p_sample(denoise_fn=denoise_fn, data=img_t, t=t_, noise_fn=noise_fn, clip_denoised=False, return_pred_xstart=False, use_var=True).detach() return img_t class PVCNN2(PVCNN2Base): sa_blocks = [ ((32, 2, 32), (1024, 0.1, 32, (32, 64))), ((64, 3, 16), (256, 0.2, 32, (64, 128))), ((128, 3, 8), (64, 0.4, 32, (128, 256))), (None, (16, 0.8, 32, (256, 256, 512))), ] fp_blocks = [ ((256, 256), (256, 3, 8)), ((256, 256), (256, 3, 8)), ((256, 128), (128, 2, 16)), ((128, 128, 64), (64, 2, 32)), ] def __init__(self, num_classes, embed_dim, use_att,dropout, extra_feature_channels=3, width_multiplier=1, voxel_resolution_multiplier=1): super().__init__( num_classes=num_classes, embed_dim=embed_dim, use_att=use_att, dropout=dropout, extra_feature_channels=extra_feature_channels, width_multiplier=width_multiplier, voxel_resolution_multiplier=voxel_resolution_multiplier ) class Model(nn.Module): def __init__(self, args, betas, loss_type: str, model_mean_type: str, model_var_type:str): super(Model, self).__init__() self.diffusion = GaussianDiffusion(betas, loss_type, model_mean_type, model_var_type) self.model = PVCNN2(num_classes=args.nc, embed_dim=args.embed_dim, use_att=args.attention, dropout=args.dropout, extra_feature_channels=0) def prior_kl(self, x0): return self.diffusion._prior_bpd(x0) def all_kl(self, x0, clip_denoised=True): total_bpd_b, vals_bt, prior_bpd_b, mse_bt = self.diffusion.calc_bpd_loop(self._denoise, x0, clip_denoised) return { 'total_bpd_b': total_bpd_b, 'terms_bpd': vals_bt, 'prior_bpd_b': prior_bpd_b, 'mse_bt':mse_bt } def _denoise(self, data, t): B, D,N= data.shape assert data.dtype == torch.float assert t.shape == torch.Size([B]) and t.dtype == torch.int64 out = self.model(data, t) assert out.shape == torch.Size([B, D, N]) return out def get_loss_iter(self, data, noises=None): B, D, N = data.shape t = torch.randint(0, self.diffusion.num_timesteps, size=(B,), device=data.device) if noises is not None: noises[t!=0] = torch.randn((t!=0).sum(), *noises.shape[1:]).to(noises) losses = self.diffusion.p_losses( denoise_fn=self._denoise, data_start=data, t=t, noise=noises) assert losses.shape == t.shape == torch.Size([B]) return losses def gen_samples(self, shape, device, noise_fn=torch.randn, constrain_fn=lambda x, t:x, clip_denoised=False, max_timestep=None, keep_running=False): return self.diffusion.p_sample_loop(self._denoise, shape=shape, device=device, noise_fn=noise_fn, constrain_fn=constrain_fn, clip_denoised=clip_denoised, max_timestep=max_timestep, keep_running=keep_running) def reconstruct(self, x0, t, constrain_fn=lambda x, t:x): return self.diffusion.reconstruct(x0, t, self._denoise, constrain_fn=constrain_fn) def train(self): self.model.train() def eval(self): self.model.eval() def multi_gpu_wrapper(self, f): self.model = f(self.model) def get_betas(schedule_type, b_start, b_end, time_num): if schedule_type == 'linear': betas = np.linspace(b_start, b_end, time_num) elif schedule_type == 'warm0.1': betas = b_end * np.ones(time_num, dtype=np.float64) warmup_time = int(time_num * 0.1) betas[:warmup_time] = np.linspace(b_start, b_end, warmup_time, dtype=np.float64) elif schedule_type == 'warm0.2': betas = b_end * np.ones(time_num, dtype=np.float64) warmup_time = int(time_num * 0.2) betas[:warmup_time] = np.linspace(b_start, b_end, warmup_time, dtype=np.float64) elif schedule_type == 'warm0.5': betas = b_end * np.ones(time_num, dtype=np.float64) warmup_time = int(time_num * 0.5) betas[:warmup_time] = np.linspace(b_start, b_end, warmup_time, dtype=np.float64) else: raise NotImplementedError(schedule_type) return betas def get_constrain_function(ground_truth, mask, eps, num_steps=1): ''' :param target_shape_constraint: target voxels :return: constrained x ''' # eps_all = list(reversed(np.linspace(0,np.float_power(eps, 1/2), 500)**2)) eps_all = list(reversed(np.linspace(0, np.sqrt(eps), 1000)**2 )) def constrain_fn(x, t): eps_ = eps_all[t] if (t<1000) else 0 for _ in range(num_steps): x = x - eps_ * ((x - ground_truth) * mask) return x return constrain_fn ############################################################################# def get_dataset(dataroot, npoints,category,use_mask=False): tr_dataset = ShapeNet15kPointClouds(root_dir=dataroot, categories=[category], split='train', tr_sample_size=npoints, te_sample_size=npoints, scale=1., normalize_per_shape=False, normalize_std_per_axis=False, random_subsample=True, use_mask = use_mask) te_dataset = ShapeNet15kPointClouds(root_dir=dataroot, categories=[category], split='val', tr_sample_size=npoints, te_sample_size=npoints, scale=1., normalize_per_shape=False, normalize_std_per_axis=False, all_points_mean=tr_dataset.all_points_mean, all_points_std=tr_dataset.all_points_std, use_mask=use_mask ) return tr_dataset, te_dataset def evaluate_gen(opt, ref_pcs, logger): if ref_pcs is None: _, test_dataset = get_dataset(opt.dataroot, opt.npoints, opt.category, use_mask=False) test_dataloader = torch.utils.data.DataLoader(test_dataset, batch_size=opt.batch_size, shuffle=False, num_workers=int(opt.workers), drop_last=False) ref = [] for data in tqdm(test_dataloader, total=len(test_dataloader), desc='Generating Samples'): x = data['test_points'] m, s = data['mean'].float(), data['std'].float() ref.append(x*s + m) ref_pcs = torch.cat(ref, dim=0).contiguous() logger.info("Loading sample path: %s" % (opt.eval_path)) sample_pcs = torch.load(opt.eval_path).contiguous() logger.info("Generation sample size:%s reference size: %s" % (sample_pcs.size(), ref_pcs.size())) # Compute metrics results = compute_all_metrics(sample_pcs, ref_pcs, opt.batch_size) results = {k: (v.cpu().detach().item() if not isinstance(v, float) else v) for k, v in results.items()} pprint(results) logger.info(results) jsd = JSD(sample_pcs.numpy(), ref_pcs.numpy()) pprint('JSD: {}'.format(jsd)) logger.info('JSD: {}'.format(jsd)) def generate(model, opt): _, test_dataset = get_dataset(opt.dataroot, opt.npoints, opt.category) test_dataloader = torch.utils.data.DataLoader(test_dataset, batch_size=opt.batch_size, shuffle=False, num_workers=int(opt.workers), drop_last=False) with torch.no_grad(): samples = [] ref = [] for i, data in tqdm(enumerate(test_dataloader), total=len(test_dataloader), desc='Generating Samples'): x = data['test_points'].transpose(1,2) m, s = data['mean'].float(), data['std'].float() gen = model.gen_samples(x.shape, 'cuda', clip_denoised=False).detach().cpu() gen = gen.transpose(1,2).contiguous() x = x.transpose(1,2).contiguous() gen = gen * s + m x = x * s + m samples.append(gen) ref.append(x) visualize_pointcloud_batch(os.path.join(str(Path(opt.eval_path).parent), 'x.png'), gen[:64], None, None, None) samples = torch.cat(samples, dim=0) ref = torch.cat(ref, dim=0) torch.save(samples, opt.eval_path) return ref def main(opt): if opt.category == 'airplane': opt.beta_start = 1e-5 opt.beta_end = 0.008 opt.schedule_type = 'warm0.1' exp_id = os.path.splitext(os.path.basename(__file__))[0] dir_id = os.path.dirname(__file__) output_dir = get_output_dir(dir_id, exp_id) copy_source(__file__, output_dir) logger = setup_logging(output_dir) outf_syn, = setup_output_subdirs(output_dir, 'syn') betas = get_betas(opt.schedule_type, opt.beta_start, opt.beta_end, opt.time_num) model = Model(opt, betas, opt.loss_type, opt.model_mean_type, opt.model_var_type) if opt.cuda: model.cuda() def _transform_(m): return nn.parallel.DataParallel(m) model = model.cuda() model.multi_gpu_wrapper(_transform_) model.eval() with torch.no_grad(): logger.info("Resume Path:%s" % opt.model) resumed_param = torch.load(opt.model) model.load_state_dict(resumed_param['model_state']) ref = None if opt.generate: opt.eval_path = os.path.join(outf_syn, 'samples.pth') Path(opt.eval_path).parent.mkdir(parents=True, exist_ok=True) ref=generate(model, opt) if opt.eval_gen: # Evaluate generation evaluate_gen(opt, ref, logger) def parse_args(): parser = argparse.ArgumentParser() parser.add_argument('--dataroot', default='/viscam/u/alexzhou907/01DATA/shapenet/ShapeNetCore.v2.PC15k', help='input batch size') parser.add_argument('--category', default='car') parser.add_argument('--batch_size', type=int, default=50, help='input batch size') parser.add_argument('--workers', type=int, default=16, help='workers') parser.add_argument('--niter', type=int, default=10000, help='number of epochs to train for') parser.add_argument('--generate',default=True) parser.add_argument('--eval_gen', default=True) parser.add_argument('--nc', default=3) parser.add_argument('--npoints', default=2048) '''model''' parser.add_argument('--beta_start', default=0.0001) parser.add_argument('--beta_end', default=0.02) parser.add_argument('--schedule_type', default='linear') parser.add_argument('--time_num', default=1000) #params parser.add_argument('--attention', default=True) parser.add_argument('--dropout', default=0.1) parser.add_argument('--embed_dim', type=int, default=64) parser.add_argument('--loss_type', default='mse') parser.add_argument('--model_mean_type', default='eps') parser.add_argument('--model_var_type', default='fixedsmall') parser.add_argument('--model', default='',required=True, help="path to model (to continue training)") '''eval''' parser.add_argument('--eval_path', default='') parser.add_argument('--manualSeed', default=42, type=int, help='random seed') parser.add_argument('--gpu', type=int, default=0, metavar='S', help='gpu id (default: 0)') opt = parser.parse_args() if torch.cuda.is_available(): opt.cuda = True else: opt.cuda = False return opt if __name__ == '__main__': opt = parse_args() set_seed(opt) main(opt) # results in /viscam/u/alexzhou907/research/diffusion/shapenet/output/test_chair