trainer_self.py

# Copyright Niantic 2019. Patent Pending. All rights reserved.
#
# This software is licensed under the terms of the Monodepth2 licence
# which allows for non-commercial use only, the full terms of which are made
# available in the LICENSE file.

from __future__ import absolute_import, division, print_function

import numpy as np
import time

import torch
import torch.nn.functional as F
import torch.optim as optim
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter

import json

from utils import *
from kitti_utils import *
from layers import *
from uncertainty_utils import *
from self_utils import SelfLoss

import datasets
import networks


class Trainer_Self:
    def __init__(self, options):
        self.opt = options
        self.log_path = os.path.join(self.opt.log_dir, self.opt.model_name)

        # checking height and width are multiples of 32
        assert self.opt.height % 32 == 0, "'height' must be a multiple of 32"
        assert self.opt.width % 32 == 0, "'width' must be a multiple of 32"

        self.models = {}
        self.teacher_models = {}
        self.parameters_to_train = []

        self.device = torch.device("cpu" if self.opt.no_cuda else "cuda")

        self.num_scales = len(self.opt.scales)
        self.num_input_frames = len(self.opt.frame_ids)
        self.num_pose_frames = 2 if self.opt.pose_model_input == "pairs" else self.num_input_frames

        assert self.opt.frame_ids[0] == 0, "frame_ids must start with 0"

        if self.opt.use_stereo:
            self.opt.frame_ids.append("s")

        self.models["encoder"] = networks.ResnetEncoder(
            self.opt.num_layers, self.opt.weights_init == "pretrained")
        self.models["encoder"].to(self.device)
        self.parameters_to_train += list(self.models["encoder"].parameters())

        self.models["depth"] = networks.DepthDecoder(
            self.models["encoder"].num_ch_enc, self.opt.scales, uncertainty=True, uncert_act=self.opt.uncert_act_stud)
        self.models["depth"].to(self.device)
        self.parameters_to_train += list(self.models["depth"].parameters())

        self.teacher_models["encoder"] = networks.ResnetEncoder(
            self.opt.num_layers, self.opt.weights_init == "pretrained")
        self.teacher_models["encoder"].to(self.device)

        self.teacher_models["depth"] = networks.DepthDecoder(
            self.teacher_models["encoder"].num_ch_enc, self.opt.scales, uncertainty=self.opt.uncertainty, uncert_act=self.opt.uncert_act)
        self.teacher_models["depth"].to(self.device)

        self.generate_depth = self.generate_depth_uncertainty
        self.generate_teacher_depth = self.generate_depth_uncertainty if self.opt.uncertainty else self.generate_depth_deterministic

        self.model_optimizer = optim.Adam(self.parameters_to_train, self.opt.learning_rate)
        self.model_lr_scheduler = optim.lr_scheduler.StepLR(
            self.model_optimizer, self.opt.scheduler_step_size, 0.1)

        if self.opt.load_weights_folder is not None:
            self.load_model()

        print("Training model named:\n  ", self.opt.model_name)
        print("Models and tensorboard events files are saved to:\n  ", self.opt.log_dir)
        print("Training is using:\n  ", self.device)

        # data
        datasets_dict = {"kitti": datasets.KITTIRAWDataset,
                         "kitti_odom": datasets.KITTIOdomDataset,
                         "scannet":datasets.ScanNetDataset}
        self.dataset = datasets_dict[self.opt.dataset]

        fpath = os.path.join(os.path.dirname(__file__), "splits", self.opt.split, "{}_files.txt")

        train_filenames = readlines(fpath.format("train"))
        val_filenames = readlines(fpath.format("val"))
        img_ext = '.png' # if self.opt.png else '.jpg'

        num_train_samples = len(train_filenames)
        self.num_total_steps = num_train_samples // self.opt.batch_size * self.opt.num_epochs

        train_dataset = self.dataset(
            self.opt.data_path, train_filenames, self.opt.height, self.opt.width,
            self.opt.frame_ids, 4, is_train=True, img_ext=img_ext)
        self.train_loader = DataLoader(
            train_dataset, self.opt.batch_size, True,
            num_workers=self.opt.num_workers, pin_memory=True, drop_last=True)
        val_dataset = self.dataset(
            self.opt.data_path, val_filenames, self.opt.height, self.opt.width,
            self.opt.frame_ids, 4, is_train=False, img_ext=img_ext)
        self.val_loader = DataLoader(
            val_dataset, self.opt.batch_size, True,
            num_workers=self.opt.num_workers, pin_memory=True, drop_last=True)
        self.val_iter = iter(self.val_loader)

        self.loss_fn = SelfLoss(stud_dist=self.opt.dist_self, 
                                stud_uncert_as_a_fraction_of_depth = self.opt.stud_uncert_as_a_fraction_of_depth, 
                                kldiv = self.opt.kldiv, 
                                teacher_dist = self.opt.distribution)

        self.writers = {}
        for mode in ["train", "val"]:
            self.writers[mode] = SummaryWriter(os.path.join(self.log_path, mode))

        self.depth_metric_names = [
            "de/abs_rel", "de/sq_rel", "de/rms", "de/log_rms", "da/a1", "da/a2", "da/a3"]

        print("Using split:\n  ", self.opt.split)
        print("There are {:d} training items and {:d} validation items\n".format(
            len(train_dataset), len(val_dataset)))

        self.save_opts()

    def set_train(self):
        """Convert all models to training mode
        """
        for m in self.models.values():
            m.train()

    def set_eval(self):
        """Convert all models to testing/evaluation mode
        """
        for m in self.models.values():
            m.eval()

    def train(self):
        """Run the entire training pipeline
        """
        self.epoch = 0
        self.step = 0
        self.start_time = time.time()
        for self.epoch in range(self.opt.num_epochs):
            self.run_epoch()
            if (self.epoch + 1) % self.opt.save_frequency == 0:
                self.save_model()

    def run_epoch(self):
        """Run a single epoch of training and validation
        """
        self.model_lr_scheduler.step()

        print("Training")
        self.set_train()

        for batch_idx, inputs in enumerate(self.train_loader):

            before_op_time = time.time()

            outputs, losses = self.process_batch(inputs)

            # if np.isnan(losses["loss"].item()):
            #     print('Current loss dict:', losses)
            #     raise ValueError("training_loss is NaN")

            self.model_optimizer.zero_grad()
            losses["loss"].backward()
            self.model_optimizer.step()

            duration = time.time() - before_op_time

            # log less frequently after the first 2000 steps to save time & disk space
            early_phase = batch_idx % self.opt.log_frequency == 0 and self.step < 2000
            late_phase = self.step % 2000 == 0

            if early_phase or late_phase:
                self.log_time(batch_idx, duration, losses["loss"].cpu().data)

                if "depth_gt" in inputs:
                    self.compute_depth_losses(inputs, outputs, losses)

                self.log("train", inputs, outputs, losses)
                self.val()

            self.step += 1

    def process_batch(self, inputs):
        """Pass a minibatch through the network and generate images and losses
        """
        for key, ipt in inputs.items():
            inputs[key] = ipt.to(self.device)

        # Otherwise, we only feed the image with frame_id 0 through the depth encoder
        features = self.models["encoder"](inputs["color_aug", 0, 0])
        outputs = self.models["depth"](features)

        with torch.no_grad():
            teacher_features = self.teacher_models["encoder"](inputs["color_aug", 0, 0])
            teacher_outputs = self.teacher_models["depth"](teacher_features)
            
        self.generate_depth(outputs, uncert_as_depth_frac=self.opt.stud_uncert_as_a_fraction_of_depth)
        self.generate_teacher_depth(teacher_outputs, uncert_as_depth_frac=self.opt.uncert_as_a_fraction_of_depth)
        losses = self.compute_losses(inputs, outputs, teacher_outputs)

        return outputs, losses


    def val(self):
        """Validate the model on a single minibatch
        """
        self.set_eval()
        try:
            inputs = next(self.val_iter)
        except StopIteration:
            self.val_iter = iter(self.val_loader)
            inputs = next(self.val_iter)

        with torch.no_grad():
            outputs, losses = self.process_batch(inputs)

            if "depth_gt" in inputs:
                self.compute_depth_losses(inputs, outputs, losses)

            self.log("val", inputs, outputs, losses)
            del inputs, outputs, losses

        self.set_train()


    def generate_depth_deterministic(self, outputs, uncert_as_depth_frac):
        """Generate the warped (reprojected) color images for a minibatch.
        Generated images are saved into the `outputs` dictionary.
        """
        for scale in self.opt.scales:
            disp = outputs[("disp", scale)]
            if not self.opt.v1_multiscale:
                disp = F.interpolate(
                    disp, [self.opt.height, self.opt.width], mode="bilinear", align_corners=False)

            _, depth = disp_to_depth(disp, self.opt.min_depth, self.opt.max_depth)

            outputs[("depth", 0, scale)] = depth


    def generate_depth_uncertainty(self, outputs, uncert_as_depth_frac):
        """Generate the warped (reprojected) color images for a minibatch.
        Generated images are saved into the `outputs` dictionary.
        """
        for scale in self.opt.scales:
            disp = outputs[("disp", scale)]
            uncert = outputs[("uncert", scale)]

            if not self.opt.v1_multiscale:
                disp = F.interpolate(
                    disp, [self.opt.height, self.opt.width], mode="bilinear", align_corners=False)
                uncert = F.interpolate(
                    uncert, [self.opt.height, self.opt.width], mode="bilinear", align_corners=False)

            _, depth = disp_to_depth(disp, self.opt.min_depth, self.opt.max_depth)

            outputs[("depth", 0, scale)] = depth
            if uncert_as_depth_frac:
                outputs[("uncert", 0, scale)] = depth * uncert
            else:
                outputs[("uncert", 0, scale)] = uncert


    def compute_losses(self, inputs, outputs, teacher_outputs):
        """Compute the reprojection and smoothness losses for a minibatch
        """
        losses = {}
        total_loss = 0

        for scale in self.opt.scales:

            if self.opt.v1_multiscale:
                source_scale = scale
            else:
                source_scale = 0
            teacher_depth = teacher_outputs[("depth", 0, source_scale)]
            pred = outputs[("depth", 0, scale)]
            teacher_uncert = teacher_outputs[("uncert", 0, source_scale)] if self.opt.uncertainty else None
            pred_uncert = outputs[("uncert", 0, scale)]
            loss = self.loss_fn(pred, pred_uncert, teacher_depth, teacher_uncert)

            if self.opt.self_scaling:
                loss /=  2 ** scale

            total_loss += loss
            losses["loss/{}".format(scale)] = loss

        total_loss /= self.num_scales
        losses["loss"] = total_loss
        return losses

    def compute_depth_losses(self, inputs, outputs, losses):
        """Compute depth metrics, to allow monitoring during training

        This isn't particularly accurate as it averages over the entire batch,
        so is only used to give an indication of validation performance
        """
        depth_pred = outputs[("depth", 0, 0)]
        depth_pred = torch.clamp(F.interpolate(
            depth_pred, [375, 1242], mode="bilinear", align_corners=False), 1e-3, 80)
        depth_pred = depth_pred.detach()

        depth_gt = inputs["depth_gt"]
        mask = depth_gt > 0

        # garg/eigen crop
        crop_mask = torch.zeros_like(mask)
        crop_mask[:, :, 153:371, 44:1197] = 1
        mask = mask * crop_mask

        depth_gt = depth_gt[mask]
        depth_pred = depth_pred[mask]
        depth_pred *= torch.median(depth_gt) / torch.median(depth_pred)

        depth_pred = torch.clamp(depth_pred, min=1e-3, max=80)

        depth_errors = compute_depth_errors(depth_gt, depth_pred)

        for i, metric in enumerate(self.depth_metric_names):
            losses[metric] = np.array(depth_errors[i].cpu())

    def log_time(self, batch_idx, duration, loss):
        """Print a logging statement to the terminal
        """
        samples_per_sec = self.opt.batch_size / duration
        time_sofar = time.time() - self.start_time
        training_time_left = (
            self.num_total_steps / self.step - 1.0) * time_sofar if self.step > 0 else 0
        print_string = "epoch {:>3} | batch {:>6} | examples/s: {:5.1f}" + \
            " | loss: {:.5f} | time elapsed: {} | time left: {}"
        print(print_string.format(self.epoch, batch_idx, samples_per_sec, loss,
                                  sec_to_hm_str(time_sofar), sec_to_hm_str(training_time_left)))

    def log(self, mode, inputs, outputs, losses):
        """Write an event to the tensorboard events file
        """
        writer = self.writers[mode]
        for l, v in losses.items():
            writer.add_scalar("{}".format(l), v, self.step)

        # for j in range(min(4, self.opt.batch_size)):  # write a maxmimum of four images
        #     for s in self.opt.scales:
        #         for frame_id in self.opt.frame_ids:
        #             writer.add_image(
        #                 "color_{}_{}/{}".format(frame_id, s, j),
        #                 inputs[("color", frame_id, s)][j].data, self.step)
        #             if s == 0 and frame_id != 0:
        #                 writer.add_image(
        #                     "color_pred_{}_{}/{}".format(frame_id, s, j),
        #                     outputs[("color", frame_id, s)][j].data, self.step)

        #         writer.add_image(
        #             "disp_{}/{}".format(s, j),
        #             normalize_image(outputs[("disp", s)][j]), self.step)

        #         if not self.opt.disable_automasking:
        #             writer.add_image(
        #                 "automask_{}/{}".format(s, j),
        #                 outputs["identity_selection/{}".format(s)][j][None, ...], self.step)

    def save_opts(self):
        """Save options to disk so we know what we ran this experiment with
        """
        models_dir = os.path.join(self.log_path, "models")
        if not os.path.exists(models_dir):
            os.makedirs(models_dir)
        to_save = self.opt.__dict__.copy()

        with open(os.path.join(models_dir, 'opt.json'), 'w') as f:
            json.dump(to_save, f, indent=2)

    def save_model(self):
        """Save model weights to disk
        """
        save_folder = os.path.join(self.log_path, "models", "weights_{}".format(self.epoch))
        if not os.path.exists(save_folder):
            os.makedirs(save_folder)

        for model_name, model in self.models.items():
            save_path = os.path.join(save_folder, "{}.pth".format(model_name))
            to_save = model.state_dict()
            if model_name == 'encoder':
                # save the sizes - these are needed at prediction time
                to_save['height'] = self.opt.height
                to_save['width'] = self.opt.width
                to_save['use_stereo'] = self.opt.use_stereo
            torch.save(to_save, save_path)

        save_path = os.path.join(save_folder, "{}.pth".format("adam"))
        torch.save(self.model_optimizer.state_dict(), save_path)

    def load_model(self):
        """Load model(s) from disk
        """
        self.opt.load_weights_folder = os.path.expanduser(self.opt.load_weights_folder)

        assert os.path.isdir(self.opt.load_weights_folder), \
            "Cannot find folder {}".format(self.opt.load_weights_folder)
        print("loading model from folder {}".format(self.opt.load_weights_folder))

        for n in self.opt.models_to_load:
            print("Loading {} as teacher weights...".format(n))
            path = os.path.join(self.opt.load_weights_folder, "{}.pth".format(n))
            model_dict = self.teacher_models[n].state_dict()
            pretrained_dict = torch.load(path)
            pretrained_dict = {k: v for k, v in pretrained_dict.items() if k in model_dict}
            model_dict.update(pretrained_dict)
            self.teacher_models[n].load_state_dict(model_dict)

        if self.opt.finetune:
            for n in self.opt.models_to_load:
                print("Loading {} as student weights...".format(n))
                path = os.path.join(self.opt.load_weights_folder, "{}.pth".format(n))
                model_dict = self.models[n].state_dict()
                pretrained_dict = torch.load(path)
                pretrained_dict = {k: v for k, v in pretrained_dict.items() if k in model_dict}
                model_dict.update(pretrained_dict)
                self.models[n].load_state_dict(model_dict)