eval.py

"""Adapted from:
    @longcw faster_rcnn_pytorch: https://github.com/longcw/faster_rcnn_pytorch
    @rbgirshick py-faster-rcnn https://github.com/rbgirshick/py-faster-rcnn
    Licensed under The MIT License [see LICENSE for details]
"""



from __future__ import print_function

import warnings
warnings.filterwarnings("ignore")

import torch
from torch.autograd import Variable
from data.config import *

import sys
import os
import time
import numpy as np
import pickle
import wandb 
from datetime import datetime
from tqdm import tqdm
from utils.visualization import predict_visualization

if sys.version_info[0] == 2:
    import xml.etree.cElementTree as ET
else:
    import xml.etree.ElementTree as ET




class Timer(object):
    """A simple timer."""
    def __init__(self):
        self.total_time = 0.
        self.calls = 0
        self.start_time = 0.
        self.diff = 0.
        self.average_time = 0.

    def tic(self):
        # using time.time instead of time.clock because time time.clock
        # does not normalize for multithreading
        self.start_time = time.time()

    def toc(self, average=True):
        self.diff = time.time() - self.start_time
        self.total_time += self.diff
        self.calls += 1
        self.average_time = self.total_time / self.calls
        if average:
            return self.average_time
        else:
            return self.diff


def parse_rec(filename):
    """ Parse a PASCAL VOC xml file """
    tree = ET.parse(filename)
    objects = []
    for obj in tree.findall('object'):
        obj_struct = {}
        obj_struct['name'] = obj.find('name').text
        # obj_struct['pose'] = obj.find('pose').text
        # obj_struct['truncated'] = int(obj.find('truncated').text)
        obj_struct['difficult'] = int(obj.find('difficult').text)
        bbox = obj.find('bndbox')
        obj_struct['bbox'] = [int(bbox.find('xmin').text) - 1,
                              int(bbox.find('ymin').text) - 1,
                              int(bbox.find('xmax').text) - 1,
                              int(bbox.find('ymax').text) - 1]
        objects.append(obj_struct)

    return objects


def get_voc_results_file_template(image_set, clss, devkit_path):
    filename = 'det_' + image_set + '_%s.txt' % (clss)
    filedir = os.path.join(devkit_path, 'results')
    if not os.path.exists(filedir):
        os.makedirs(filedir)
    path = os.path.join(filedir, filename)
    return path


def write_voc_results_file(all_boxes, dataset, labelmap, set_type, devkit_path):
    for cls_ind, clss in enumerate(labelmap):
        filename = get_voc_results_file_template(set_type, clss, devkit_path)
        with open(filename, 'wt') as f:
            for im_ind, index in enumerate(dataset.ids):
                dets = all_boxes[cls_ind+1][im_ind]
                if dets == []:
                    continue
                # the VOCdevkit expects 1-based indices
                for k in range(dets.shape[0]):
                    f.write('{:s} {:.3f} {:.1f} {:.1f} {:.1f} {:.1f}\n'.
                            format(index[1], dets[k, -1],
                                   dets[k, 0] + 1, dets[k, 1] + 1,
                                   dets[k, 2] + 1, dets[k, 3] + 1))


def do_python_eval(logger, annopath, imgsetpath, devkit_path, labelmap, set_type, output_dir='output', use_07=True):
    cachedir = os.path.join(devkit_path, 'annotations_cache')
    aps = []
    # The PASCAL VOC metric changed in 2010
    use_07_metric = use_07
    if not os.path.isdir(output_dir):
        os.mkdir(output_dir)
    for i, clss in enumerate(labelmap):
        filename = get_voc_results_file_template(set_type, clss, devkit_path)
        rec, prec, ap = voc_eval(
           filename, annopath, imgsetpath.format(set_type), clss, cachedir,
           ovthresh=0.5, use_07_metric=use_07_metric)
        aps += [ap]
        logger.info('AP for {} = {:.4f}'.format(clss, ap))
        wandb.log({f"val/{clss}": ap})
        with open(os.path.join(output_dir, clss + '_pr.pkl'), 'wb') as f:
            pickle.dump({'rec': rec, 'prec': prec, 'ap': ap}, f)
    logger.info('Mean AP = {:.4f}'.format(np.mean(aps)))
    wandb.log({f"val/mAP": np.mean(aps)})
    logger.info('~~~~~~~~')




def voc_ap(rec, prec, use_07_metric=True):
    """ ap = voc_ap(rec, prec, [use_07_metric])
    Compute VOC AP given precision and recall.
    If use_07_metric is true, uses the
    VOC 07 11 point method (default:True).
    """
    if use_07_metric:
        # 11 point metric
        ap = 0.
        for t in np.arange(0., 1.1, 0.1):
            if np.sum(rec >= t) == 0:
                p = 0
            else:
                p = np.max(prec[rec >= t])
            ap = ap + p / 11.
    else:
        # correct AP calculation
        # first append sentinel values at the end
        mrec = np.concatenate(([0.], rec, [1.]))
        mpre = np.concatenate(([0.], prec, [0.]))

        # compute the precision envelope
        for i in range(mpre.size - 1, 0, -1):
            mpre[i - 1] = np.maximum(mpre[i - 1], mpre[i])

        # to calculate area under PR curve, look for points
        # where X axis (recall) changes value
        i = np.where(mrec[1:] != mrec[:-1])[0]

        # and sum (\Delta recall) * prec
        ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])
    return ap


def voc_eval(detpath,
             annopath,
             imagesetfile,
             classname,
             cachedir,
             ovthresh=0.5,
             use_07_metric=True):
    """rec, prec, ap = voc_eval(detpath,
                           annopath,
                           imagesetfile,
                           classname,
                           [ovthresh],
                           [use_07_metric])
Top level function that does the PASCAL VOC evaluation.
detpath: Path to detections
   detpath.format(classname) should produce the detection results file.
annopath: Path to annotations
   annopath.format(imagename) should be the xml annotations file.
imagesetfile: Text file containing the list of images, one image per line.
classname: Category name (duh)
cachedir: Directory for caching the annotations
[ovthresh]: Overlap threshold (default = 0.5)
[use_07_metric]: Whether to use VOC07's 11 point AP computation
   (default True)
"""

    if not os.path.isdir(cachedir):
        os.mkdir(cachedir)
    cachefile = os.path.join(cachedir, 'annots.pkl')
    # read list of images
    with open(imagesetfile, 'r') as f:
        lines = f.readlines()
    imagenames = [x.strip() for x in lines]
    if not os.path.isfile(cachefile):
        # load annots
        recs = {}
        for i, imagename in enumerate(imagenames):
            recs[imagename] = parse_rec(annopath % (imagename))
            
        # save
        with open(cachefile, 'wb') as f:
            pickle.dump(recs, f)
    else:
        # load
        with open(cachefile, 'rb') as f:
            recs = pickle.load(f)

    # extract gt objects for this class
    class_recs = {}
    npos = 0
    for imagename in imagenames:
        R = [obj for obj in recs[imagename] if obj['name'] == classname]
        bbox = np.array([x['bbox'] for x in R])
        difficult = np.array([x['difficult'] for x in R]).astype(bool)
        det = [False] * len(R)
        npos = npos + sum(~difficult)
        class_recs[imagename] = {'bbox': bbox,
                                 'difficult': difficult,
                                 'det': det}

    # read dets
    detfile = detpath.format(classname)
    with open(detfile, 'r') as f:
        lines = f.readlines()
    if any(lines) == 1:
        splitlines = [x.strip().split(' ') for x in lines]
        image_ids = [x[0] for x in splitlines]
        confidence = np.array([float(x[1]) for x in splitlines])
        BB = np.array([[float(z) for z in x[2:]] for x in splitlines])

        # sort by confidence
        sorted_ind = np.argsort(-confidence)
        sorted_scores = np.sort(-confidence)
        BB = BB[sorted_ind, :]
        image_ids = [image_ids[x] for x in sorted_ind]

        # go down dets and mark TPs and FPs
        nd = len(image_ids)
        tp = np.zeros(nd)
        fp = np.zeros(nd)
        for d in range(nd):
            R = class_recs[image_ids[d]]
            bb = BB[d, :].astype(float)
            ovmax = -np.inf
            BBGT = R['bbox'].astype(float)
            if BBGT.size > 0:
                # compute overlaps
                # intersection
                ixmin = np.maximum(BBGT[:, 0], bb[0])
                iymin = np.maximum(BBGT[:, 1], bb[1])
                ixmax = np.minimum(BBGT[:, 2], bb[2])
                iymax = np.minimum(BBGT[:, 3], bb[3])
                iw = np.maximum(ixmax - ixmin, 0.)
                ih = np.maximum(iymax - iymin, 0.)
                inters = iw * ih
                uni = ((bb[2] - bb[0]) * (bb[3] - bb[1]) +
                       (BBGT[:, 2] - BBGT[:, 0]) *
                       (BBGT[:, 3] - BBGT[:, 1]) - inters)
                overlaps = inters / uni
                ovmax = np.max(overlaps)
                jmax = np.argmax(overlaps)

            if ovmax > ovthresh:
                if not R['difficult'][jmax]:
                    if not R['det'][jmax]:
                        tp[d] = 1.
                        R['det'][jmax] = 1
                    else:
                        fp[d] = 1.
            else:
                fp[d] = 1.

        # compute precision recall
        fp = np.cumsum(fp)
        tp = np.cumsum(tp)
        rec = tp / float(npos)
        # avoid divide by zero in case the first detection matches a difficult
        # ground truth
        prec = tp / np.maximum(tp + fp, np.finfo(np.float64).eps)
        ap = voc_ap(rec, prec, use_07_metric)
    else:
        rec = -1.
        prec = -1.
        ap = -1.
    
    return rec, prec, ap


def test_net(annopath, imgsetpath, labelmap, save_folder, net, cuda, dataset, transform, top_k,
             im_size=300, thresh=0.05, phase='test', set_type=None, devkit_path=None, logger=None):
    num_images = len(dataset)

    all_boxes = [[[] for _ in range(num_images)]
                 for _ in range(len(labelmap)+1)]

    # timers
    _t = {'im_detect': Timer(), 'misc': Timer()}
    output_dir = save_folder

    if not os.path.exists(output_dir):
        os.makedirs(output_dir)

    det_file = os.path.join(output_dir, 'detections.pkl')

    date = f"{datetime.now().hour}-{datetime.now().minute}"
    
    test_iterator = tqdm(range(num_images),
                          desc="Validating... (loss=X.X)",
                          bar_format="{l_bar}{r_bar}",
                          dynamic_ncols=False,
                          )
    
    tsne_tensors = []
    for i, _ in enumerate(test_iterator):
        im, gt, h, w = dataset.pull_item(i)

        x = Variable(im.unsqueeze(0))
        if cuda:
            x = x.cuda()
        _t['im_detect'].tic()
        detections, feats  = net(x, phase)
        detections =detections.data
        detect_time = _t['im_detect'].toc(average=False)
      
        tsne_tensors.append(feats[-1].detach())

        pred = []
        # skip j = 0, because it's the background class
        for j in range(1, detections.size(1)):
            dets = detections[0, j, :]
            mask = dets[:, 0].gt(0.).expand(5, dets.size(0)).t()
            dets = torch.masked_select(dets, mask).view(-1, 5)
            if dets.size(0) == 0:
                continue
            boxes = dets[:, 1:]
            boxes[:, 0] *= w
            boxes[:, 2] *= w
            boxes[:, 1] *= h
            boxes[:, 3] *= h
            scores = dets[:, 0].cpu().numpy()
            cls_dets = np.hstack((boxes.cpu().numpy(),
                                  scores[:, np.newaxis])).astype(np.float32,
                                                                 copy=False)
            all_boxes[j][i] = cls_dets

            result = [boxes[i] for i in range(len(scores)) if scores[i] >= thresh]
            if len(result) != 0:
                pred.append(result)
        
        test_iterator.set_description("Detection... (im_detect= %.2d/%.2d:%.3fs)" % (i+1, num_images, detect_time))

        if i % 50 == 1:
            pred = torch.cat([p[0].unsqueeze(0) for p in pred], dim=0)
            res = predict_visualization(im, gt, pred)
            mode = f"_{datetime.now().hour}-{datetime.now().minute}"
            images = wandb.Image(res, caption=mode)
            wandb.log({"Predictions": images})
            # featmap_visualization(im, feats, layer=0, iter=f'{date}_{i}')
            # actmap_visualization(im, net, feats)


    tsne_cat = torch.cat(tsne_tensors, dim=0).squeeze(-1).squeeze(-1)

    with open(det_file, 'wb') as f:
        pickle.dump(all_boxes, f, pickle.HIGHEST_PROTOCOL)

    evaluate_detections(all_boxes, output_dir, dataset, labelmap, set_type, devkit_path, annopath, imgsetpath, logger)

    return tsne_cat


def evaluate_detections(box_list, output_dir, dataset, labelmap, set_type, devkit_path, annopath, imgsetpath, logger):
    write_voc_results_file(box_list, dataset, labelmap, set_type, devkit_path)
    do_python_eval(logger, annopath, imgsetpath, devkit_path, labelmap, set_type, output_dir)