import os import random from datetime import datetime from typing import Union, Optional, Tuple, List, Dict, Set import numpy as np import torch import torch.nn as nn import torch.nn.functional as F import torch.optim as optim from torch.utils.data import DataLoader from torch.utils.data.dataloader import default_collate from torch.utils.tensorboard import SummaryWriter from tqdm import tqdm from models.rgb_part_net import RGBPartNet from utils.dataset import CASIAB, ClipConditions, ClipViews, ClipClasses from utils.sampler import TripletSampler from utils.triplet_loss import JointBatchTripletLoss, BatchTripletLoss class Model: def __init__( self, system_config: Dict, model_config: Dict, hyperparameter_config: Dict ): self.disable_acc = system_config.get('disable_acc', False) if self.disable_acc: self.device = torch.device('cpu') else: # Enable accelerator if torch.cuda.is_available(): self.device = torch.device('cuda') else: print('No accelerator available, fallback to CPU.') self.device = torch.device('cpu') self.save_dir = system_config.get('save_dir', 'runs') if not os.path.exists(self.save_dir): os.makedirs(self.save_dir) self.checkpoint_dir = os.path.join(self.save_dir, 'checkpoint') self.log_dir = os.path.join(self.save_dir, 'logs') for dir_ in (self.log_dir, self.checkpoint_dir): if not os.path.exists(dir_): os.mkdir(dir_) self.meta = model_config self.hp = hyperparameter_config self.curr_iter = self.meta.get('restore_iter', 0) self.total_iter = self.meta.get('total_iter', 80_000) self.curr_iters = self.meta.get('restore_iters', (0, 0, 0)) self.total_iters = self.meta.get('total_iters', (80000, 80000, 80000)) self.is_train: bool = True self.in_channels: int = 3 self.in_size: Tuple[int, int] = (64, 48) self.pr: Optional[int] = None self.k: Optional[int] = None self._gallery_dataset_meta: Optional[Dict[str, List]] = None self._probe_datasets_meta: Optional[Dict[str, Dict[str, List]]] = None self._model_name: str = self.meta.get('name', 'RGB-GaitPart') self._hp_sig: str = self._make_signature(self.hp) self._dataset_sig: str = 'undefined' self.rgb_pn: Optional[RGBPartNet] = None self.triplet_loss: Optional[JointBatchTripletLoss] = None self.optimizer: Optional[optim.Adam] = None self.scheduler: Optional[optim.lr_scheduler.StepLR] = None self.writer: Optional[SummaryWriter] = None self.image_log_on = system_config.get('image_log_on', False) self.CASIAB_GALLERY_SELECTOR = { 'selector': {'conditions': ClipConditions({r'nm-0[1-4]'})} } self.CASIAB_PROBE_SELECTORS = { 'nm': {'selector': {'conditions': ClipConditions({r'nm-0[5-6]'})}}, 'bg': {'selector': {'conditions': ClipConditions({r'bg-0[1-2]'})}}, 'cl': {'selector': {'conditions': ClipConditions({r'cl-0[1-2]'})}}, } @property def _model_sig(self) -> str: return '_'.join( (self._model_name, str(self.curr_iter), str(self.total_iter)) ) @property def _checkpoint_sig(self) -> str: return '_'.join((self._model_sig, self._hp_sig, self._dataset_sig, str(self.pr), str(self.k))) @property def _checkpoint_name(self) -> str: return os.path.join(self.checkpoint_dir, self._checkpoint_sig) @property def _log_sig(self) -> str: return '_'.join((self._model_name, str(self.total_iter), self._hp_sig, self._dataset_sig, str(self.pr), str(self.k))) @property def _log_name(self) -> str: return os.path.join(self.log_dir, self._log_sig) def fit_all( self, dataset_config: Dict, dataset_selectors: Dict[ str, Dict[str, Union[ClipClasses, ClipConditions, ClipViews]] ], dataloader_config: Dict, ): for (curr_iter, total_iter, (condition, selector)) in zip( self.curr_iters, self.total_iters, dataset_selectors.items() ): print(f'Training model {condition} ...') # Skip finished model if curr_iter == total_iter: continue # Check invalid restore iter elif curr_iter > total_iter: raise ValueError("Restore iter '{}' should less than total " "iter '{}'".format(curr_iter, total_iter)) self.curr_iter = curr_iter self.total_iter = total_iter self.fit( dict(**dataset_config, **{'selector': selector}), dataloader_config ) def fit( self, dataset_config: Dict, dataloader_config: Dict, ): self.is_train = True dataset = self._parse_dataset_config(dataset_config) dataloader = self._parse_dataloader_config(dataset, dataloader_config) # Prepare for model, optimizer and scheduler model_hp: Dict = self.hp.get('model', {}).copy() triplet_is_hard = model_hp.pop('triplet_is_hard', True) triplet_is_mean = model_hp.pop('triplet_is_mean', True) triplet_margins = model_hp.pop('triplet_margins', None) optim_hp: Dict = self.hp.get('optimizer', {}).copy() ae_optim_hp = optim_hp.pop('auto_encoder', {}) pn_optim_hp = optim_hp.pop('part_net', {}) hpm_optim_hp = optim_hp.pop('hpm', {}) fc_optim_hp = optim_hp.pop('fc', {}) sched_hp = self.hp.get('scheduler', {}) self.rgb_pn = RGBPartNet(self.in_channels, self.in_size, **model_hp, image_log_on=self.image_log_on) # Hard margins if triplet_margins: # Same margins if triplet_margins[0] == triplet_margins[1]: self.triplet_loss = BatchTripletLoss( triplet_is_hard, triplet_margins[0] ) else: # Different margins self.triplet_loss = JointBatchTripletLoss( self.rgb_pn.hpm_num_parts, triplet_is_hard, triplet_is_mean, triplet_margins ) else: # Soft margins self.triplet_loss = BatchTripletLoss( triplet_is_hard, triplet_is_mean, None ) num_pairs = (self.pr*self.k-1) * (self.pr*self.k) // 2 num_pos_pairs = (self.k*(self.k-1)//2) * self.pr # Try to accelerate computation using CUDA or others self.rgb_pn = self.rgb_pn.to(self.device) self.triplet_loss = self.triplet_loss.to(self.device) self.optimizer = optim.Adam([ {'params': self.rgb_pn.ae.parameters(), **ae_optim_hp}, {'params': self.rgb_pn.pn.parameters(), **pn_optim_hp}, {'params': self.rgb_pn.hpm.parameters(), **hpm_optim_hp}, {'params': self.rgb_pn.fc_mat, **fc_optim_hp} ], **optim_hp) sched_final_gamma = sched_hp.get('final_gamma', 0.001) sched_start_step = sched_hp.get('start_step', 15_000) all_step = self.total_iter - sched_start_step def lr_lambda(epoch): if epoch > sched_start_step: passed_step = epoch - sched_start_step return sched_final_gamma ** (passed_step / all_step) else: return 1 self.scheduler = optim.lr_scheduler.LambdaLR(self.optimizer, lr_lambda=[ lr_lambda, lr_lambda, lr_lambda, lr_lambda ]) self.writer = SummaryWriter(self._log_name) # Set seeds for reproducibility random.seed(0) torch.manual_seed(0) self.rgb_pn.train() # Init weights at first iter if self.curr_iter == 0: self.rgb_pn.apply(self.init_weights) else: # Load saved state dicts checkpoint = torch.load(self._checkpoint_name) random.setstate(checkpoint['rand_states'][0]) torch.set_rng_state(checkpoint['rand_states'][1]) self.rgb_pn.load_state_dict(checkpoint['model_state_dict']) self.optimizer.load_state_dict(checkpoint['optim_state_dict']) self.scheduler.load_state_dict(checkpoint['sched_state_dict']) # Training start start_time = datetime.now() running_loss = torch.zeros(5, device=self.device) print(f"{'Time':^8} {'Iter':^5} {'Loss':^6}", f"{'Xrecon':^8} {'CanoCons':^8} {'PoseSim':^8}", f"{'BATripH':^8} {'BATripP':^8} {'LR':^9}") for (batch_c1, batch_c2) in dataloader: self.curr_iter += 1 # Zero the parameter gradients self.optimizer.zero_grad() # forward + backward + optimize x_c1 = batch_c1['clip'].to(self.device) x_c2 = batch_c2['clip'].to(self.device) embedding, ae_losses, images = self.rgb_pn(x_c1, x_c2) y = batch_c1['label'].to(self.device) # Duplicate labels for each part y = y.repeat(self.rgb_pn.num_total_parts, 1) trip_loss, dist, num_non_zero = self.triplet_loss(embedding, y) losses = torch.cat(( ae_losses, torch.stack(( trip_loss[:self.rgb_pn.hpm_num_parts].mean(), trip_loss[self.rgb_pn.hpm_num_parts:].mean() )) )) loss = losses.sum() loss.backward() self.optimizer.step() # Statistics and checkpoint running_loss += losses.detach() # Write losses to TensorBoard self.writer.add_scalar('Loss/all', loss, self.curr_iter) self.writer.add_scalars('Loss/disentanglement', dict(zip(( 'Cross reconstruction loss', 'Canonical consistency loss', 'Pose similarity loss' ), ae_losses)), self.curr_iter) self.writer.add_scalars('Loss/triplet loss', { 'HPM': losses[3], 'PartNet': losses[4] }, self.curr_iter) # None-zero losses in batch if num_non_zero is not None: self.writer.add_scalars('Loss/non-zero counts', { 'HPM': num_non_zero[:self.rgb_pn.hpm_num_parts].mean(), 'PartNet': num_non_zero[self.rgb_pn.hpm_num_parts:].mean() }, self.curr_iter) # Embedding distance mean_hpm_dist = dist[:self.rgb_pn.hpm_num_parts].mean(0) self._add_ranked_scalars( 'Embedding/HPM distance', mean_hpm_dist, num_pos_pairs, num_pairs, self.curr_iter ) mean_pa_dist = dist[self.rgb_pn.hpm_num_parts:].mean(0) self._add_ranked_scalars( 'Embedding/ParNet distance', mean_pa_dist, num_pos_pairs, num_pairs, self.curr_iter ) # Embedding norm mean_hpm_embedding = embedding[:self.rgb_pn.hpm_num_parts].mean(0) mean_hpm_norm = mean_hpm_embedding.norm(dim=-1) self._add_ranked_scalars( 'Embedding/HPM norm', mean_hpm_norm, self.k, self.pr * self.k, self.curr_iter ) mean_pa_embedding = embedding[self.rgb_pn.hpm_num_parts:].mean(0) mean_pa_norm = mean_pa_embedding.norm(dim=-1) self._add_ranked_scalars( 'Embedding/PartNet norm', mean_pa_norm, self.k, self.pr * self.k, self.curr_iter ) # Learning rate lrs = self.scheduler.get_last_lr() # Write learning rates self.writer.add_scalar( 'Learning rate', lrs[0], self.curr_iter ) if self.curr_iter % 100 == 0: # Write disentangled images if self.image_log_on: i_a, i_c, i_p = images self.writer.add_images( 'Appearance image', i_a, self.curr_iter ) self.writer.add_images( 'Canonical image', i_c, self.curr_iter ) for i, (o, p) in enumerate(zip(x_c1, i_p)): self.writer.add_images( f'Original image/batch {i}', o, self.curr_iter ) self.writer.add_images( f'Pose image/batch {i}', p, self.curr_iter ) time_used = datetime.now() - start_time remaining_minute, second = divmod(time_used.seconds, 60) hour, minute = divmod(remaining_minute, 60) print(f'{hour:02}:{minute:02}:{second:02}', f'{self.curr_iter:5d} {running_loss.sum() / 100:6.3f}', '{:f} {:f} {:f} {:f} {:f}'.format(*running_loss / 100), f'{lrs[0]:.3e}') running_loss.zero_() # Step scheduler self.scheduler.step() if self.curr_iter % 1000 == 0: torch.save({ 'rand_states': (random.getstate(), torch.get_rng_state()), 'model_state_dict': self.rgb_pn.state_dict(), 'optim_state_dict': self.optimizer.state_dict(), 'sched_state_dict': self.scheduler.state_dict(), }, self._checkpoint_name) if self.curr_iter == self.total_iter: self.writer.close() break def _add_ranked_scalars( self, main_tag: str, metric: torch.Tensor, num_pos: int, num_all: int, global_step: int ): rank = metric.argsort() pos_ile = 100 - (num_pos - 1) * 100 // num_all self.writer.add_scalars(main_tag, { '0%-ile': metric[rank[-1]], f'{100 - pos_ile}%-ile': metric[rank[-num_pos]], '50%-ile': metric[rank[num_all // 2 - 1]], f'{pos_ile}%-ile': metric[rank[num_pos - 1]], '100%-ile': metric[rank[0]] }, global_step) def predict_all( self, iters: Tuple[int], dataset_config: Dict, dataset_selectors: Dict[ str, Dict[str, Union[ClipClasses, ClipConditions, ClipViews]] ], dataloader_config: Dict, ) -> Dict[str, torch.Tensor]: # Transform data to features gallery_samples, probe_samples = self.transform( iters, dataset_config, dataset_selectors, dataloader_config ) # Evaluate features accuracy = self.evaluate(gallery_samples, probe_samples) return accuracy def transform( self, iters: Tuple[int], dataset_config: Dict, dataset_selectors: Dict[ str, Dict[str, Union[ClipClasses, ClipConditions, ClipViews]] ], dataloader_config: Dict ): self.is_train = False # Split gallery and probe dataset gallery_dataloader, probe_dataloaders = self._split_gallery_probe( dataset_config, dataloader_config ) # Get pretrained models at iter_ checkpoints = self._load_pretrained( iters, dataset_config, dataset_selectors ) # Init models model_hp: dict = self.hp.get('model', {}).copy() model_hp.pop('triplet_is_hard', True) model_hp.pop('triplet_is_mean', True) model_hp.pop('triplet_margins', None) self.rgb_pn = RGBPartNet(self.in_channels, self.in_size, **model_hp) # Try to accelerate computation using CUDA or others self.rgb_pn = self.rgb_pn.to(self.device) self.rgb_pn.eval() gallery_samples, probe_samples = [], {} # Gallery checkpoint = torch.load(list(checkpoints.values())[0]) self.rgb_pn.load_state_dict(checkpoint['model_state_dict']) for sample in tqdm(gallery_dataloader, desc='Transforming gallery', unit='clips'): gallery_samples.append(self._get_eval_sample(sample)) gallery_samples = default_collate(gallery_samples) # Probe for (condition, dataloader) in probe_dataloaders.items(): checkpoint = torch.load(checkpoints[condition]) self.rgb_pn.load_state_dict(checkpoint['model_state_dict']) probe_samples_c = [] for sample in tqdm(dataloader, desc=f'Transforming probe {condition}', unit='clips'): probe_samples_c.append(self._get_eval_sample(sample)) probe_samples[condition] = default_collate(probe_samples_c) return gallery_samples, probe_samples def _get_eval_sample(self, sample: Dict[str, Union[List, torch.Tensor]]): label = sample.pop('label').item() clip = sample.pop('clip').to(self.device) with torch.no_grad(): feature = self.rgb_pn(clip) return { **{'label': label}, **sample, **{'feature': feature} } def evaluate( self, gallery_samples: Dict[str, Union[List[str], torch.Tensor]], probe_samples: Dict[str, Dict[str, Union[List[str], torch.Tensor]]], num_ranks: int = 5 ) -> Dict[str, torch.Tensor]: probe_conditions = self._probe_datasets_meta.keys() gallery_views_meta = self._gallery_dataset_meta['views'] probe_views_meta = list(self._probe_datasets_meta.values())[0]['views'] accuracy = { condition: torch.empty( len(gallery_views_meta), len(probe_views_meta), num_ranks ) for condition in self._probe_datasets_meta.keys() } (labels_g, _, views_g, features_g) = gallery_samples.values() views_g = np.asarray(views_g) for (v_g_i, view_g) in enumerate(gallery_views_meta): gallery_view_mask = (views_g == view_g) f_g = features_g[gallery_view_mask] y_g = labels_g[gallery_view_mask] for condition in probe_conditions: probe_samples_c = probe_samples[condition] accuracy_c = accuracy[condition] (labels_p, _, views_p, features_p) = probe_samples_c.values() views_p = np.asarray(views_p) for (v_p_i, view_p) in enumerate(probe_views_meta): probe_view_mask = (views_p == view_p) f_p = features_p[probe_view_mask] y_p = labels_p[probe_view_mask] # Euclidean distance f_p_squared_sum = torch.sum(f_p ** 2, dim=1).unsqueeze(1) f_g_squared_sum = torch.sum(f_g ** 2, dim=1).unsqueeze(0) f_p_times_f_g_sum = f_p @ f_g.T dist = torch.sqrt(F.relu( f_p_squared_sum - 2*f_p_times_f_g_sum + f_g_squared_sum )) # Ranked accuracy rank_mask = dist.argsort(1)[:, :num_ranks] positive_mat = torch.eq(y_p.unsqueeze(1), y_g[rank_mask]).cumsum(1).gt(0) positive_counts = positive_mat.sum(0) total_counts, _ = dist.size() accuracy_c[v_g_i, v_p_i, :] = positive_counts / total_counts return accuracy def _load_pretrained( self, iters: Tuple[int], dataset_config: Dict, dataset_selectors: Dict[ str, Dict[str, Union[ClipClasses, ClipConditions, ClipViews]] ] ) -> Dict[str, str]: checkpoints = {} for (iter_, (condition, selector)) in zip( iters, dataset_selectors.items() ): self.curr_iter = iter_ self._dataset_sig = self._make_signature( dict(**dataset_config, **selector), popped_keys=['root_dir', 'cache_on'] ) checkpoints[condition] = self._checkpoint_name return checkpoints def _split_gallery_probe( self, dataset_config: Dict, dataloader_config: Dict, ) -> Tuple[DataLoader, Dict[str, DataLoader]]: dataset_name = dataset_config.get('name', 'CASIA-B') if dataset_name == 'CASIA-B': gallery_dataset = self._parse_dataset_config( dict(**dataset_config, **self.CASIAB_GALLERY_SELECTOR) ) self._gallery_dataset_meta = gallery_dataset.metadata gallery_dataloader = self._parse_dataloader_config( gallery_dataset, dataloader_config ) probe_datasets = { condition: self._parse_dataset_config( dict(**dataset_config, **selector) ) for (condition, selector) in self.CASIAB_PROBE_SELECTORS.items() } self._probe_datasets_meta = { condition: dataset.metadata for (condition, dataset) in probe_datasets.items() } probe_dataloaders = { condition: self._parse_dataloader_config( dataset, dataloader_config ) for (condition, dataset) in probe_datasets.items() } elif dataset_name == 'FVG': # TODO gallery_dataloader = None probe_dataloaders = None else: raise ValueError('Invalid dataset: {0}'.format(dataset_name)) return gallery_dataloader, probe_dataloaders @staticmethod def init_weights(m): if isinstance(m, nn.modules.conv._ConvNd): nn.init.normal_(m.weight, 0.0, 0.01) elif isinstance(m, nn.modules.batchnorm._NormBase): nn.init.normal_(m.weight, 1.0, 0.01) nn.init.zeros_(m.bias) elif isinstance(m, nn.Linear): nn.init.xavier_uniform_(m.weight) elif isinstance(m, RGBPartNet): nn.init.xavier_uniform_(m.fc_mat) def _parse_dataset_config( self, dataset_config: Dict ) -> Union[CASIAB]: self.in_channels = dataset_config.get('num_input_channels', 3) self.in_size = dataset_config.get('frame_size', (64, 48)) self._dataset_sig = self._make_signature( dataset_config, popped_keys=['root_dir', 'cache_on'] ) config: Dict = dataset_config.copy() name = config.pop('name', 'CASIA-B') if name == 'CASIA-B': return CASIAB(**config, is_train=self.is_train) elif name == 'FVG': # TODO pass raise ValueError('Invalid dataset: {0}'.format(name)) def _parse_dataloader_config( self, dataset: Union[CASIAB], dataloader_config: Dict ) -> DataLoader: config: Dict = dataloader_config.copy() (self.pr, self.k) = config.pop('batch_size', (8, 16)) if self.is_train: triplet_sampler = TripletSampler(dataset, (self.pr, self.k)) return DataLoader(dataset, batch_sampler=triplet_sampler, collate_fn=self._batch_splitter, **config) else: # is_test return DataLoader(dataset, **config) def _batch_splitter( self, batch: List[Dict[str, Union[np.int64, str, torch.Tensor]]] ) -> Tuple[Dict[str, Union[List[str], torch.Tensor]], Dict[str, Union[List[str], torch.Tensor]]]: """ Disentanglement need two random conditions, this function will split pr * k * 2 samples to 2 dicts each containing pr * k samples. labels and clip data are tensor, and others are list. """ _batch = [[], []] for i in range(0, self.pr * self.k * 2, self.k * 2): _batch[0] += batch[i:i + self.k] _batch[1] += batch[i + self.k:i + self.k * 2] return default_collate(_batch[0]), default_collate(_batch[1]) def _make_signature(self, config: Dict, popped_keys: Optional[List] = None) -> str: _config = config.copy() if popped_keys: for key in popped_keys: _config.pop(key, None) return self._gen_sig(list(_config.values())) def _gen_sig(self, values: Union[Tuple, List, Set, str, int, float]) -> str: strings = [] for v in values: if isinstance(v, str): strings.append(v) elif isinstance(v, (Tuple, List)): strings.append(self._gen_sig(v)) elif isinstance(v, Set): strings.append(self._gen_sig(sorted(list(v)))) elif isinstance(v, Dict): strings.append(self._gen_sig(list(v.values()))) else: strings.append(str(v)) return '_'.join(strings)