utils.py

import numpy as np
import subprocess
import os


def ioa_with_anchors(anchors_min, anchors_max, box_min, box_max):
    # calculate the overlap proportion between the anchor and all bbox for supervise signal,
    # the length of the anchor is 0.01
    len_anchors = anchors_max - anchors_min
    int_xmin = np.maximum(anchors_min, box_min)
    int_xmax = np.minimum(anchors_max, box_max)
    inter_len = np.maximum(int_xmax - int_xmin, 0.)
    scores = np.divide(inter_len, len_anchors)
    return scores


def iou_with_anchors(anchors_min, anchors_max, box_min, box_max):
    """Compute jaccard score between a box and the anchors.
    """
    len_anchors = anchors_max - anchors_min
    int_xmin = np.maximum(anchors_min, box_min)
    int_xmax = np.minimum(anchors_max, box_max)
    inter_len = np.maximum(int_xmax - int_xmin, 0.)
    union_len = len_anchors - inter_len + box_max - box_min
    # print inter_len,union_len
    jaccard = np.divide(inter_len, union_len)
    return jaccard


class ProposalGenerator(object):
    def __init__(self, temporal_dim=None, max_duration=None, annotations=None):
        self.tscale = temporal_dim
        self.max_duration = max_duration
        self.annots = annotations  # For THUMOS only
        self.rescale_segment = self.rescale_segment_anet if self.annots is None else self.rescale_segment_thumos

    def rescale_segment_anet(self, start_index, end_index, video_name=None):
        return start_index / self.tscale, end_index / self.tscale

    def rescale_segment_thumos(self, start_index, end_index, video_name=None):
        b = self.annots[video_name]['start_snippet']
        d = self.annots[video_name]['master_snippet_duration']
        return (start_index + b) / d, (end_index + b) / d

    def __call__(self, start, end, confidence_map, video_names):
        batch_props = []
        for i, video_name in enumerate(video_names):
            start_scores = start[i]
            end_scores = end[i]
            clr_confidence = (confidence_map[i][1])
            reg_confidence = (confidence_map[i][0])

            max_start = max(start_scores)
            max_end = max(end_scores)

            # generate the set of start points and end points
            start_bins = np.zeros(self.tscale)
            start_bins[0] = 1  # [1,0,0...,0,1]
            for idx in range(1, self.tscale - 1):
                if start_scores[idx] > start_scores[idx + 1] and start_scores[idx] > start_scores[idx - 1]:
                    start_bins[idx] = 1
                elif start_scores[idx] > (0.5 * max_start):
                    start_bins[idx] = 1

            end_bins = np.zeros(len(end_scores))
            end_bins[-1] = 1
            for idx in range(1, self.tscale - 1):
                if end_scores[idx] > end_scores[idx + 1] and end_scores[idx] > end_scores[idx - 1]:
                    end_bins[idx] = 1
                elif end_scores[idx] > (0.5 * max_end):
                    end_bins[idx] = 1

            # generate proposals
            new_props = []
            for idx in range(self.max_duration):
                for jdx in range(self.tscale):
                    start_index = jdx
                    end_index = start_index + idx + 1
                    if end_index < self.tscale and start_bins[start_index] == 1 and end_bins[end_index] == 1:
                        xmin, xmax = self.rescale_segment(start_index, end_index, video_name)
                        xmin_score = start_scores[start_index]
                        xmax_score = end_scores[end_index]
                        clr_score = clr_confidence[idx, jdx]
                        reg_score = reg_confidence[idx, jdx]
                        score = xmin_score * xmax_score * clr_score * reg_score
                        new_props.append([xmin, xmax, xmin_score, xmax_score, clr_score, reg_score, score])
            new_props = np.stack(new_props)
            batch_props.append(new_props)

        return batch_props