From 1373b363b5369b1938c5deafdf19e36ec6c3f6d4 Mon Sep 17 00:00:00 2001 From: Hunor Laczko Date: Wed, 15 Mar 2023 22:20:33 +0100 Subject: [PATCH] activate all levels --- cvae_pyramid.py | 186 ++++++++++++++++-------------- models/pyramid_model_optimized.py | 2 +- 2 files changed, 99 insertions(+), 89 deletions(-) diff --git a/cvae_pyramid.py b/cvae_pyramid.py index 1fa2632..c4b647c 100644 --- a/cvae_pyramid.py +++ b/cvae_pyramid.py @@ -222,65 +222,70 @@ def train_step(batch): out_levels = model([levels_input, conds], training=True) if config['use_skinning']: - mesh_orig, uv_orig = batch['mesh_unposed'], batch['uv_unposed'] + mesh_orig = batch['mesh_unposed'] else: - mesh_orig, uv_orig = batch['mesh_posed'], batch['uv_posed'] + mesh_orig = batch['mesh_posed'] output = dict() reconstr_loss_total = 0 kl_total = 0 - #for i, res in enumerate(config['pyramid']['resolutions']): - i = -1 - res = config['pyramid']['resolutions'][-1] - res_modifier = '_' + str(res) if res != 512 else '' - pred, mean, logvar = out_levels[i] + for i, res in enumerate(config['pyramid']['resolutions']): + # i = -1 + res = config['pyramid']['resolutions'][i] if i != len(config['pyramid']['resolutions']) - 1 else 512 + res_modifier = '_' + str(res) if res != 512 else '' + pred, mean, logvar = out_levels[i] - mesh_pred = uv2mesh(pred) + if config['use_skinning']: + uv_orig = batch['uv_unposed' + res_modifier] + else: + uv_orig = batch['uv_posed' + res_modifier] - if config['use_skinning']: - shape = batch['pose_shape'][:, 72:] - gender = tf.cast(tf.greater_equal(shape[:, 10] , 0.5), tf.int32) - shape_offsets = tf.stack([tf.gather(tf.tensordot(shape[i,:10][tf.newaxis, :], smpl.shapedirs[gender[i]], axes=[[1], [2]])[0,:,:], smpl.to_keep, axis=0) for i in range(shape.shape[0])]) + mesh_pred = uv2mesh(pred) - mesh_pred_shaped = mesh_pred + shape_offsets * batch['mesh_mask'][:,:,tf.newaxis] - mesh_pred_posed = smpl.run_garment(batch['pose_shape'][:, :72], batch['pose_shape'][:, 72:], mesh_pred_shaped) - else: - mesh_pred_posed = mesh_pred + if config['use_skinning']: + shape = batch['pose_shape'][:, 72:] + gender = tf.cast(tf.greater_equal(shape[:, 10] , 0.5), tf.int32) + shape_offsets = tf.stack([tf.gather(tf.tensordot(shape[i,:10][tf.newaxis, :], smpl.shapedirs[gender[i]], axes=[[1], [2]])[0,:,:], smpl.to_keep, axis=0) for i in range(shape.shape[0])]) - # is the output already masked? + mesh_pred_shaped = mesh_pred + shape_offsets * batch['mesh_mask'][:,:,tf.newaxis] + mesh_pred_posed = smpl.run_garment(batch['pose_shape'][:, :72], batch['pose_shape'][:, 72:], mesh_pred_shaped) + else: + mesh_pred_posed = mesh_pred - e_loss = tf.reduce_mean(edge_loss(mesh_pred, mesh_orig, batch['mesh_mask'])) - c_loss, c_ratio = collision_penalty(mesh_pred_posed, batch['mesh_body_posed'], batch['mesh_mask']) - c_loss, c_ratio = tf.reduce_mean(c_loss), tf.reduce_mean(c_ratio) - l2_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(((uv_orig - pred) ** 2), axis=-1), axis=[1, 2])) - # ms_ssim = tf.reduce_mean(tf.image.ssim_multiscale(pred + 5, batch['uv_unposed'] + 5, max_val=10, power_factors=(0.5, 0.2, 0.1, 0.1, 0.1))) - # ms_ssim = tf.reduce_mean(tf.image.ssim(pred + 1, batch['uv_unposed'] + 1, max_val=3)) - l2_mesh_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (batch['mesh_posed'] - mesh_pred_posed) ** 2, axis=-1), axis=-1)) - n_loss = normal_loss(mesh_pred_posed, batch['mesh_posed'], batch['mesh_mask'], is_skirt=config['is_skirt']) - kl = kl_loss(mean, logvar) + # is the output already masked? - if config['use_extra_losses']: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss + l2_mesh_loss) - if not config['is_cape']: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (e_loss + c_loss) * config['reg_loss_coeff'] - else: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss) + e_loss = tf.reduce_mean(edge_loss(mesh_pred, mesh_orig, batch['mesh_mask'])) + c_loss, c_ratio = collision_penalty(mesh_pred_posed, batch['mesh_body_posed'], batch['mesh_mask']) + c_loss, c_ratio = tf.reduce_mean(c_loss), tf.reduce_mean(c_ratio) + l2_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(((uv_orig - pred) ** 2), axis=-1) * batch['uv_mask' + res_modifier], axis=[1, 2])) + # ms_ssim = tf.reduce_mean(tf.image.ssim_multiscale(pred + 5, batch['uv_unposed'] + 5, max_val=10, power_factors=(0.5, 0.2, 0.1, 0.1, 0.1))) + # ms_ssim = tf.reduce_mean(tf.image.ssim(pred + 1, batch['uv_unposed'] + 1, max_val=3)) + l2_mesh_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (batch['mesh_posed'] - mesh_pred_posed) ** 2, axis=-1), axis=-1)) + n_loss = normal_loss(mesh_pred_posed, batch['mesh_posed'], batch['mesh_mask'], is_skirt=config['is_skirt']) + kl = kl_loss(mean, logvar) - if config['use_norm_loss'] and step > 2000: - reconstr_loss_total = reconstr_loss_total + n_loss * config['reg_loss_coeff'] - - kl_total += kl + if config['use_extra_losses']: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss + l2_mesh_loss) + if not config['is_cape']: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (e_loss + c_loss) * config['reg_loss_coeff'] + else: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss) - mean = tf.reduce_mean(mean) - logvar = tf.reduce_mean(logvar) + if config['use_norm_loss'] and step > 2000: + reconstr_loss_total = reconstr_loss_total + n_loss * config['reg_loss_coeff'] + + kl_total += kl - metric = tf.reduce_sum(tf.math.sqrt(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (mesh_orig - mesh_pred) ** 2, axis=-1))) / tf.reduce_sum(batch['mesh_mask']) + mean = tf.reduce_mean(mean) + logvar = tf.reduce_mean(logvar) - output = { **output, **{ 'train/edge_loss' + res_modifier: e_loss, 'train/collision_loss' + res_modifier: c_loss,'train/euclid_error' + res_modifier: metric, - 'train/mean' + res_modifier: mean, 'train/logvar' + res_modifier: logvar, 'train/kl_loss' + res_modifier: kl, - 'train/var' + res_modifier: np.e ** logvar, 'train/l2_loss' + res_modifier: l2_loss, 'train/normal_loss': n_loss, - 'train/euclid_error': metric, 'train/l2_mesh_loss': l2_mesh_loss } } # 'train/ms_ssim' + res_modifier: ms_ssim, + metric = tf.reduce_sum(tf.math.sqrt(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (mesh_orig - mesh_pred) ** 2, axis=-1))) / tf.reduce_sum(batch['mesh_mask']) + + output = { **output, **{ 'train/edge_loss' + res_modifier: e_loss, 'train/collision_loss' + res_modifier: c_loss,'train/euclid_error' + res_modifier: metric, + 'train/mean' + res_modifier: mean, 'train/logvar' + res_modifier: logvar, 'train/kl_loss' + res_modifier: kl, + 'train/var' + res_modifier: np.e ** logvar, 'train/l2_loss' + res_modifier: l2_loss, 'train/normal_loss': n_loss, + 'train/euclid_error': metric, 'train/l2_mesh_loss': l2_mesh_loss } } # 'train/ms_ssim' + res_modifier: ms_ssim, loss = reconstr_loss_total + balancing_coeff * kl_total @@ -296,64 +301,69 @@ def valid_step(batch): if config['use_skinning']: - mesh_orig, uv_orig = batch['mesh_unposed'], batch['uv_unposed'] + mesh_orig = batch['mesh_unposed'] else: - mesh_orig, uv_orig = batch['mesh_posed'], batch['uv_posed'] + mesh_orig = batch['mesh_posed'] output = dict() reconstr_loss_total = 0 kl_total = 0 - #for i, res in enumerate(config['pyramid']['resolutions']): - i = -1 - res = config['pyramid']['resolutions'][-1] - res_modifier = '_' + str(res) if res != 512 else '' - pred, mean, logvar = out_levels[i] + for i, res in enumerate(config['pyramid']['resolutions']): + # i = -1 + res = config['pyramid']['resolutions'][i] if i != len(config['pyramid']['resolutions']) - 1 else 512 + res_modifier = '_' + str(res) if res != 512 else '' + pred, mean, logvar = out_levels[i] - mesh_pred = uv2mesh(pred) + if config['use_skinning']: + uv_orig = batch['uv_unposed' + res_modifier] + else: + uv_orig = batch['uv_posed' + res_modifier] - if config['use_skinning']: - shape = batch['pose_shape'][:, 72:] - gender = tf.cast(tf.greater_equal(shape[:, 10] , 0.5), tf.int32) - shape_offsets = tf.stack([tf.gather(tf.tensordot(shape[i,:10][tf.newaxis, :], smpl.shapedirs[gender[i]], axes=[[1], [2]])[0,:,:], smpl.to_keep, axis=0) for i in range(shape.shape[0])]) + mesh_pred = uv2mesh(pred) - mesh_pred_shaped = mesh_pred + shape_offsets * batch['mesh_mask'][:,:,tf.newaxis] - mesh_pred_posed = smpl.run_garment(batch['pose_shape'][:, :72], batch['pose_shape'][:, 72:], mesh_pred_shaped) - else: - mesh_pred_posed = mesh_pred - # is the output already masked? + if config['use_skinning']: + shape = batch['pose_shape'][:, 72:] + gender = tf.cast(tf.greater_equal(shape[:, 10] , 0.5), tf.int32) + shape_offsets = tf.stack([tf.gather(tf.tensordot(shape[i,:10][tf.newaxis, :], smpl.shapedirs[gender[i]], axes=[[1], [2]])[0,:,:], smpl.to_keep, axis=0) for i in range(shape.shape[0])]) - e_loss = tf.reduce_mean(edge_loss(mesh_pred, mesh_orig, batch['mesh_mask'])) - c_loss, c_ratio = collision_penalty(mesh_pred_posed, batch['mesh_body_posed'], batch['mesh_mask']) - c_loss, c_ratio = tf.reduce_mean(c_loss), tf.reduce_mean(c_ratio) - l2_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(((uv_orig - pred) ** 2), axis=-1), axis=[1, 2])) - # ms_ssim = tf.reduce_mean(tf.image.ssim_multiscale(pred + 5, batch['uv_unposed'] + 5, max_val=10, power_factors=(0.5, 0.2, 0.1, 0.1, 0.1))) - # ms_ssim = tf.reduce_mean(tf.image.ssim(pred + 1, batch['uv_unposed'] + 1, max_val=3)) - l2_mesh_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (batch['mesh_posed'] - mesh_pred_posed) ** 2, axis=-1), axis=-1)) - n_loss = normal_loss(mesh_pred_posed, batch['mesh_posed'], batch['mesh_mask'], is_skirt=config['is_skirt']) - kl = kl_loss(mean, logvar) + mesh_pred_shaped = mesh_pred + shape_offsets * batch['mesh_mask'][:,:,tf.newaxis] + mesh_pred_posed = smpl.run_garment(batch['pose_shape'][:, :72], batch['pose_shape'][:, 72:], mesh_pred_shaped) + else: + mesh_pred_posed = mesh_pred + # is the output already masked? - if config['use_extra_losses']: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss + l2_mesh_loss) - if not config['is_cape']: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (e_loss + c_loss) * config['reg_loss_coeff'] - else: - reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss) - - if config['use_norm_loss'] and step > 2000: - reconstr_loss_total = reconstr_loss_total + n_loss * config['reg_loss_coeff'] + e_loss = tf.reduce_mean(edge_loss(mesh_pred, mesh_orig, batch['mesh_mask'])) + c_loss, c_ratio = collision_penalty(mesh_pred_posed, batch['mesh_body_posed'], batch['mesh_mask']) + c_loss, c_ratio = tf.reduce_mean(c_loss), tf.reduce_mean(c_ratio) + l2_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(((uv_orig - pred) ** 2), axis=-1) * batch['uv_mask' + res_modifier], axis=[1, 2])) + # ms_ssim = tf.reduce_mean(tf.image.ssim_multiscale(pred + 5, batch['uv_unposed'] + 5, max_val=10, power_factors=(0.5, 0.2, 0.1, 0.1, 0.1))) + # ms_ssim = tf.reduce_mean(tf.image.ssim(pred + 1, batch['uv_unposed'] + 1, max_val=3)) + l2_mesh_loss = tf.reduce_mean(1e3 * tf.reduce_mean(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (batch['mesh_posed'] - mesh_pred_posed) ** 2, axis=-1), axis=-1)) + n_loss = normal_loss(mesh_pred_posed, batch['mesh_posed'], batch['mesh_mask'], is_skirt=config['is_skirt']) + kl = kl_loss(mean, logvar) + + if config['use_extra_losses']: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss + l2_mesh_loss) + if not config['is_cape']: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (e_loss + c_loss) * config['reg_loss_coeff'] + else: + reconstr_loss_total = reconstr_loss_total + config['pyramid']['loss_weights'][i] * (l2_loss) - kl_total += kl + if config['use_norm_loss'] and step > 2000: + reconstr_loss_total = reconstr_loss_total + n_loss * config['reg_loss_coeff'] + + kl_total += kl - mean = tf.reduce_mean(mean) - logvar = tf.reduce_mean(logvar) + mean = tf.reduce_mean(mean) + logvar = tf.reduce_mean(logvar) - metric = tf.reduce_sum(tf.math.sqrt(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (mesh_orig - mesh_pred) ** 2, axis=-1))) / tf.reduce_sum(batch['mesh_mask']) + metric = tf.reduce_sum(tf.math.sqrt(tf.reduce_sum(batch['mesh_mask'][:,:,tf.newaxis] * (mesh_orig - mesh_pred) ** 2, axis=-1))) / tf.reduce_sum(batch['mesh_mask']) - output = { **output, **{ 'valid/edge_loss' + res_modifier: e_loss, 'valid/collision_loss' + res_modifier: c_loss,'valid/euclid_error' + res_modifier: metric, - 'valid/mean' + res_modifier: mean, 'valid/logvar' + res_modifier: logvar, 'valid/kl_loss' + res_modifier: kl, - 'valid/var' + res_modifier: np.e ** logvar, 'valid/l2_loss' + res_modifier: l2_loss, 'valid/normal_loss': n_loss, - 'valid/euclid_error': metric, 'valid/l2_mesh_loss': l2_mesh_loss } } # 'valid/ms_ssim' + res_modifier: ms_ssim, + output = { **output, **{ 'valid/edge_loss' + res_modifier: e_loss, 'valid/collision_loss' + res_modifier: c_loss,'valid/euclid_error' + res_modifier: metric, + 'valid/mean' + res_modifier: mean, 'valid/logvar' + res_modifier: logvar, 'valid/kl_loss' + res_modifier: kl, + 'valid/var' + res_modifier: np.e ** logvar, 'valid/l2_loss' + res_modifier: l2_loss, 'valid/normal_loss': n_loss, + 'valid/euclid_error': metric, 'valid/l2_mesh_loss': l2_mesh_loss } } # 'valid/ms_ssim' + res_modifier: ms_ssim, loss = reconstr_loss_total + balancing_coeff * kl_total @@ -871,8 +881,8 @@ if __name__ == '__main__': # load_model(model.upscale_enhancer, config['pyramid']['final_upscale_enhancer']) # model.upscale_enhancer.trainable = False - for i in range(len(model.levels)-1): - model.levels[i].trainable = False + # for i in range(len(model.levels)-1): + # model.levels[i].trainable = False # load_model(model.cond_encoder, config['pyramid']['pretrained_pyramid']) # load_model(model.levels[0], config['pyramid']['pretrained_pyramid']) diff --git a/models/pyramid_model_optimized.py b/models/pyramid_model_optimized.py index 5019c7a..1a750b8 100644 --- a/models/pyramid_model_optimized.py +++ b/models/pyramid_model_optimized.py @@ -416,8 +416,8 @@ class PyramidNew(tf.keras.Model): pred = pred + levels_input[0][2] # uv_body_posed_32 if 1 != len(self.levels): pred = self.levels[0].enhance(pred, levels_input[0][-2], training=training) # levels_input[i][-2]=uv_mask_res - pred = tf.image.resize(pred, (self.levels[1].size, self.levels[1].size//2), tf.image.ResizeMethod.BILINEAR) out_levels.append([pred, mean, logvar]) + pred = tf.image.resize(pred, (self.levels[1].size, self.levels[1].size//2), tf.image.ResizeMethod.BILINEAR) pred = pred * levels_input[1][-2][:,:,:,tf.newaxis] else: # pred = self.enhance(pred, levels_input[0][-2], training=training)