import hashlib import pprint import tqdm import cv2 import re import json import io import glob import xarray import numpy import json import glob import io import os import pandas import pickle import subprocess def kernel_1(): t4 = 'kernel_1-t3.dat' def preprocess(t4): t1 = '/kaggle/input/mlb-player-digital-engagement-forecasting' t2 = glob.glob( os.path.join( t1, '*.csv' ) ) t3 = { o : pandas.read_csv(o).to_xarray() for o in t2 } with io.open(t4, 'wb') as f: pickle.dump(t3, f) if not os.path.exists(t4): preprocess(t4=t4) with io.open(t4, 'rb') as f: t3 = pickle.load(f) return dict( t3=t3, ) def kernel_2( o_1=None, ): t1 = {} for k in [ 'playerTwitterFollowers', 'teamTwitterFollowers', 'games', 'events' ]: t4 = '%s.nc' % k if not os.path.exists(t4): print('started %s' % t4) t2 = '/kaggle/input/mlb-player-digital-engagement-forecasting/train.csv' t3 = pandas.DataFrame( sum( [ json.loads(o) for o in o_1['t3'][t2][k].values if isinstance(o, str) ], [] ) ).to_xarray() t3.to_netcdf(t4) print('cached %s' % t4) if k == 'events': t5 = '%s-v2.nc' % k if not os.path.exists(t5): t2 = xarray.load_dataset(t4) t3 = t2.sel( index=numpy.arange( 2017653 - 10 * 1000, 2017653 + 1 ) ) t3.to_netcdf(t5) t1[k] = xarray.load_dataset(t5) print('loaded %s' % t5) else: t1[k] = xarray.load_dataset(t4) print('loaded %s' % t4) return dict( t1=t1, ) def kernel_3(should_exist=None): if should_exist is None: should_exist = False t3 = [ ('playerTwitterFollowers', None), ('teamTwitterFollowers', None), ('games', None), ('events', 'events-v2.nc'), ] o_1 = None o_2 = None t4 = '/kaggle/input/garbage' t5 = {} for k, v in t3: if v is None: t1 = os.path.join( t4, '%s.nc' % k, ) else: t1 = os.path.join( t4, v, ) if os.path.exists(t1): t2 = xarray.load_dataset(t1) else: if should_exist: pprint.pprint([k, v, t1]) raise NotImplementedError if o_1 is None: o_1 = kernel_1() if o_2 is None: o_2 = kernel_2( o_1=o_1 ) t2 = o_2['events'] t5[k] = t2 return dict( t5=t5, ) def kernel_4( o_3=None, ): [ print( o_3['t5']['events'].to_dataframe().iloc[k].to_json(indent=4) ) for k in range(-10, -1) ] [ print( o_3['t5']['games'].to_dataframe().iloc[k].to_json(indent=4) ) for k in range(-10, -1) ] t4 = 'https://www.youtube.com/watch?v=reaC7BHgL3M' r""" { "gamePk":634280, "gameType":"R", "season":2021, "gameDate":"2021-04-30", "gameTimeUTC":"2021-04-30T23:37:00Z", "resumeDate":"", "resumedFrom":"", "codedGameState":"F", "detailedGameState":"Final", "isTie":0.0, "gameNumber":1, "doubleHeader":"N", "dayNight":"night", "scheduledInnings":9, "gamesInSeries":3.0, "seriesDescription":"Regular Season", "homeId":141, "homeName":"Toronto Blue Jays", "homeAbbrev":"TOR", "homeWins":12, "homeLosses":12, "homeWinPct":0.5, "homeWinner":true, "homeScore":13.0, "awayId":144, "awayName":"Atlanta Braves", "awayAbbrev":"ATL", "awayWins":12.0, "awayLosses":14.0, "awayWinPct":0.462, "awayWinner":false, "awayScore":5.0 } """ t1 = numpy.where(o_3['t5']['events']['gamePk'] == 634280)[0] t5 = o_3['t5']['events'].index.data t6 = t5[t1] t2 = o_3['t5']['events'].sel(index=t6) t3 = o_3['t5']['games'].to_dataframe().iloc[-2].to_dict() pprint.pprint(t3) assert t3['gamePk'] == 634280 t7 = 'https://www.youtube.com/watch?v=T0MUK91ZWys' return dict( t2=t2, t3=t3, t4=t4, t7=t7, ) def kernel_5(o_4): for o in [o_4['t4'], o_4['t7']]: subprocess.check_call( [ 'youtube-dl', '-f', '18', o, ] ) def kernel_12(): import easyocr t6 = easyocr.Reader(['en']) return dict( t6=t6, ) def kernel_6( o_7=None, o_10=None, o_12=None, max_frames=None, ): if max_frames is None: max_frames = 10 import tqdm import cv2 t1 = glob.glob('*.mp4') t8 = [] for o in t1: t7 = [] t2 = None try: t2 = cv2.VideoCapture(o) for k in tqdm.tqdm(range(max_frames)): t3 = t2.read() assert t3[0] t4 = t3[1] if not o_12 is None: t5 = o_12['t6'].readtext(t4) else: t5 = None if not o_7 is None: t10 = o_7['estimate_pose'](t4) else: t10 = None if not o_10 is None: t11 = o_10['model'](t4).pandas().xywhn else: t11 = None t7.append( dict( frame_id=k, t5=t5, t10=t10, t11=t11, ), ) finally: if not t2 is None: t2.release() t8.append( dict( video_path=o, frames=t7, ) ) t9 = [] for o in t1: cap = None try: cap = cv2.VideoCapture(o) fps = cap.get(cv2.CAP_PROP_FPS) # OpenCV2 version 2 used "CV_CAP_PROP_FPS" frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT)) duration = frame_count/fps finally: if not cap is None: cap.release() t9.append( dict( video_path=o, fps=fps, frame_count=frame_count, duration=duration, ) ) return dict( t8=t8, t9=t9, ) def kernel_7( use_gpu=None, ): #!/usr/bin/env python # coding: utf-8 # # # NOTE: Turn on Internet and GPU # The code hidden below handles all the imports and function definitions (the heavy lifting). If you're a beginner I'd advice you skip this for now. When you are able to understand the rest of the code, come back here and understand each function to get a deeper knowledge. # In[1]: # !/usr/bin/env python3 # coding=utf-8 # author=dave.fang@outlook.com # create=20171225 import os import pprint import cv2 import sys import math import time import tempfile import numpy as np import matplotlib.pyplot as plt import torch import torch.nn as nn import torch.nn.parallel import torch.backends.cudnn as cudnn import torch.optim as optim import torchvision.transforms as transforms import torchvision.datasets as datasets import torchvision.models as models from torch.autograd import Variable from scipy.ndimage.filters import gaussian_filter #get_ipython().run_line_magic('matplotlib', 'inline') #get_ipython().run_line_magic('config', "InlineBackend.figure_format = 'retina'") # find connection in the specified sequence, center 29 is in the position 15 limb_seq = [[2, 3], [2, 6], [3, 4], [4, 5], [6, 7], [7, 8], [2, 9], [9, 10], [10, 11], [2, 12], [12, 13], [13, 14], [2, 1], [1, 15], [15, 17], [1, 16], [16, 18], [3, 17], [6, 18]] # the middle joints heatmap correpondence map_ids = [[31, 32], [39, 40], [33, 34], [35, 36], [41, 42], [43, 44], [19, 20], [21, 22], [23, 24], [25, 26], [27, 28], [29, 30], [47, 48], [49, 50], [53, 54], [51, 52], [55, 56], [37, 38], [45, 46]] # these are the colours for the 18 body points colors = [[255, 0, 0], [255, 85, 0], [255, 170, 0], [255, 255, 0], [170, 255, 0], [85, 255, 0], [0, 255, 0], [0, 255, 85], [0, 255, 170], [0, 255, 255], [0, 170, 255], [0, 85, 255], [0, 0, 255], [85, 0, 255], [170, 0, 255], [255, 0, 255], [255, 0, 170], [255, 0, 85]] class PoseEstimation(nn.Module): def __init__(self, model_dict): super(PoseEstimation, self).__init__() self.model0 = model_dict['block_0'] self.model1_1 = model_dict['block1_1'] self.model2_1 = model_dict['block2_1'] self.model3_1 = model_dict['block3_1'] self.model4_1 = model_dict['block4_1'] self.model5_1 = model_dict['block5_1'] self.model6_1 = model_dict['block6_1'] self.model1_2 = model_dict['block1_2'] self.model2_2 = model_dict['block2_2'] self.model3_2 = model_dict['block3_2'] self.model4_2 = model_dict['block4_2'] self.model5_2 = model_dict['block5_2'] self.model6_2 = model_dict['block6_2'] def forward(self, x): out1 = self.model0(x) out1_1 = self.model1_1(out1) out1_2 = self.model1_2(out1) out2 = torch.cat([out1_1, out1_2, out1], 1) out2_1 = self.model2_1(out2) out2_2 = self.model2_2(out2) out3 = torch.cat([out2_1, out2_2, out1], 1) out3_1 = self.model3_1(out3) out3_2 = self.model3_2(out3) out4 = torch.cat([out3_1, out3_2, out1], 1) out4_1 = self.model4_1(out4) out4_2 = self.model4_2(out4) out5 = torch.cat([out4_1, out4_2, out1], 1) out5_1 = self.model5_1(out5) out5_2 = self.model5_2(out5) out6 = torch.cat([out5_1, out5_2, out1], 1) out6_1 = self.model6_1(out6) out6_2 = self.model6_2(out6) return out6_1, out6_2 def make_layers(layer_dict): layers = [] for i in range(len(layer_dict) - 1): layer = layer_dict[i] for k in layer: v = layer[k] if 'pool' in k: layers += [nn.MaxPool2d(kernel_size=v[0], stride=v[1], padding=v[2])] else: conv2d = nn.Conv2d(in_channels=v[0], out_channels=v[1], kernel_size=v[2], stride=v[3], padding=v[4]) layers += [conv2d, nn.ReLU(inplace=True)] layer = list(layer_dict[-1].keys()) k = layer[0] v = layer_dict[-1][k] conv2d = nn.Conv2d(in_channels=v[0], out_channels=v[1], kernel_size=v[2], stride=v[3], padding=v[4]) layers += [conv2d] return nn.Sequential(*layers) def get_pose_model(): blocks = {} block_0 = [{'conv1_1': [3, 64, 3, 1, 1]}, {'conv1_2': [64, 64, 3, 1, 1]}, {'pool1_stage1': [2, 2, 0]}, {'conv2_1': [64, 128, 3, 1, 1]}, {'conv2_2': [128, 128, 3, 1, 1]}, {'pool2_stage1': [2, 2, 0]}, {'conv3_1': [128, 256, 3, 1, 1]}, {'conv3_2': [256, 256, 3, 1, 1]}, {'conv3_3': [256, 256, 3, 1, 1]}, {'conv3_4': [256, 256, 3, 1, 1]}, {'pool3_stage1': [2, 2, 0]}, {'conv4_1': [256, 512, 3, 1, 1]}, {'conv4_2': [512, 512, 3, 1, 1]}, {'conv4_3_CPM': [512, 256, 3, 1, 1]}, {'conv4_4_CPM': [256, 128, 3, 1, 1]}] blocks['block1_1'] = [{'conv5_1_CPM_L1': [128, 128, 3, 1, 1]}, {'conv5_2_CPM_L1': [128, 128, 3, 1, 1]}, {'conv5_3_CPM_L1': [128, 128, 3, 1, 1]}, {'conv5_4_CPM_L1': [128, 512, 1, 1, 0]}, {'conv5_5_CPM_L1': [512, 38, 1, 1, 0]}] blocks['block1_2'] = [{'conv5_1_CPM_L2': [128, 128, 3, 1, 1]}, {'conv5_2_CPM_L2': [128, 128, 3, 1, 1]}, {'conv5_3_CPM_L2': [128, 128, 3, 1, 1]}, {'conv5_4_CPM_L2': [128, 512, 1, 1, 0]}, {'conv5_5_CPM_L2': [512, 19, 1, 1, 0]}] for i in range(2, 7): blocks['block%d_1' % i] = [{'Mconv1_stage%d_L1' % i: [185, 128, 7, 1, 3]}, {'Mconv2_stage%d_L1' % i: [128, 128, 7, 1, 3]}, {'Mconv3_stage%d_L1' % i: [128, 128, 7, 1, 3]}, {'Mconv4_stage%d_L1' % i: [128, 128, 7, 1, 3]}, {'Mconv5_stage%d_L1' % i: [128, 128, 7, 1, 3]}, {'Mconv6_stage%d_L1' % i: [128, 128, 1, 1, 0]}, {'Mconv7_stage%d_L1' % i: [128, 38, 1, 1, 0]}] blocks['block%d_2' % i] = [{'Mconv1_stage%d_L2' % i: [185, 128, 7, 1, 3]}, {'Mconv2_stage%d_L2' % i: [128, 128, 7, 1, 3]}, {'Mconv3_stage%d_L2' % i: [128, 128, 7, 1, 3]}, {'Mconv4_stage%d_L2' % i: [128, 128, 7, 1, 3]}, {'Mconv5_stage%d_L2' % i: [128, 128, 7, 1, 3]}, {'Mconv6_stage%d_L2' % i: [128, 128, 1, 1, 0]}, {'Mconv7_stage%d_L2' % i: [128, 19, 1, 1, 0]}] layers = [] for block in block_0: # print(block) for key in block: v = block[key] if 'pool' in key: layers += [nn.MaxPool2d(kernel_size=v[0], stride=v[1], padding=v[2])] else: conv2d = nn.Conv2d(in_channels=v[0], out_channels=v[1], kernel_size=v[2], stride=v[3], padding=v[4]) layers += [conv2d, nn.ReLU(inplace=True)] models = { 'block_0': nn.Sequential(*layers) } for k in blocks: v = blocks[k] models[k] = make_layers(v) return PoseEstimation(models) def get_paf_and_heatmap(model, img_raw, scale_search, param_stride=8, box_size=368): multiplier = [scale * box_size / img_raw.shape[0] for scale in scale_search] heatmap_avg = torch.zeros((len(multiplier), 19, img_raw.shape[0], img_raw.shape[1])).cuda() paf_avg = torch.zeros((len(multiplier), 38, img_raw.shape[0], img_raw.shape[1])).cuda() for i, scale in enumerate(multiplier): img_test = cv2.resize(img_raw, (0, 0), fx=scale, fy=scale, interpolation=cv2.INTER_CUBIC) img_test_pad, pad = pad_right_down_corner(img_test, param_stride, param_stride) img_test_pad = np.transpose(np.float32(img_test_pad[:, :, :, np.newaxis]), (3, 2, 0, 1)) / 256 - 0.5 feed = Variable(torch.from_numpy(img_test_pad)).cuda() output1, output2 = model(feed) #print(output1.size()) #print(output2.size()) heatmap = nn.UpsamplingBilinear2d((img_raw.shape[0], img_raw.shape[1])).cuda()(output2) paf = nn.UpsamplingBilinear2d((img_raw.shape[0], img_raw.shape[1])).cuda()(output1) heatmap_avg[i] = heatmap[0].data paf_avg[i] = paf[0].data heatmap_avg = torch.transpose(torch.transpose(torch.squeeze(torch.mean(heatmap_avg, 0)), 0, 1), 1, 2).cuda() heatmap_avg = heatmap_avg.cpu().numpy() paf_avg = torch.transpose(torch.transpose(torch.squeeze(torch.mean(paf_avg, 0)), 0, 1), 1, 2).cuda() paf_avg = paf_avg.cpu().numpy() return paf_avg, heatmap_avg def extract_heatmap_info(heatmap_avg, param_thre1=0.1): all_peaks = [] peak_counter = 0 for part in range(18): map_ori = heatmap_avg[:, :, part] map_gau = gaussian_filter(map_ori, sigma=3) map_left = np.zeros(map_gau.shape) map_left[1:, :] = map_gau[:-1, :] map_right = np.zeros(map_gau.shape) map_right[:-1, :] = map_gau[1:, :] map_up = np.zeros(map_gau.shape) map_up[:, 1:] = map_gau[:, :-1] map_down = np.zeros(map_gau.shape) map_down[:, :-1] = map_gau[:, 1:] peaks_binary = np.logical_and.reduce( (map_gau >= map_left, map_gau >= map_right, map_gau >= map_up, map_gau >= map_down, map_gau > param_thre1)) peaks = zip(np.nonzero(peaks_binary)[1], np.nonzero(peaks_binary)[0]) # note reverse peaks = list(peaks) peaks_with_score = [x + (map_ori[x[1], x[0]],) for x in peaks] ids = range(peak_counter, peak_counter + len(peaks)) peaks_with_score_and_id = [peaks_with_score[i] + (ids[i],) for i in range(len(ids))] all_peaks.append(peaks_with_score_and_id) peak_counter += len(peaks) return all_peaks def extract_paf_info(img_raw, paf_avg, all_peaks, param_thre2=0.05, param_thre3=0.5): connection_all = [] special_k = [] mid_num = 10 for k in range(len(map_ids)): score_mid = paf_avg[:, :, [x - 19 for x in map_ids[k]]] candA = all_peaks[limb_seq[k][0] - 1] candB = all_peaks[limb_seq[k][1] - 1] nA = len(candA) nB = len(candB) if nA != 0 and nB != 0: connection_candidate = [] for i in range(nA): for j in range(nB): vec = np.subtract(candB[j][:2], candA[i][:2]) norm = math.sqrt(vec[0] * vec[0] + vec[1] * vec[1]) if norm < 1e-8: raise ZeroDivisionError vec = np.divide(vec, norm) startend = zip(np.linspace(candA[i][0], candB[j][0], num=mid_num), np.linspace(candA[i][1], candB[j][1], num=mid_num)) startend = list(startend) vec_x = np.array([score_mid[int(round(startend[I][1])), int(round(startend[I][0])), 0] for I in range(len(startend))]) vec_y = np.array([score_mid[int(round(startend[I][1])), int(round(startend[I][0])), 1] for I in range(len(startend))]) score_midpts = np.multiply(vec_x, vec[0]) + np.multiply(vec_y, vec[1]) score_with_dist_prior = sum(score_midpts) / len(score_midpts) score_with_dist_prior += min(0.5 * img_raw.shape[0] / norm - 1, 0) criterion1 = len(np.nonzero(score_midpts > param_thre2)[0]) > 0.8 * len(score_midpts) criterion2 = score_with_dist_prior > 0 if criterion1 and criterion2: connection_candidate.append( [i, j, score_with_dist_prior, score_with_dist_prior + candA[i][2] + candB[j][2]]) connection_candidate = sorted(connection_candidate, key=lambda x: x[2], reverse=True) connection = np.zeros((0, 5)) for c in range(len(connection_candidate)): i, j, s = connection_candidate[c][0:3] if i not in connection[:, 3] and j not in connection[:, 4]: connection = np.vstack([connection, [candA[i][3], candB[j][3], s, i, j]]) if len(connection) >= min(nA, nB): break connection_all.append(connection) else: special_k.append(k) connection_all.append([]) return special_k, connection_all def get_subsets(connection_all, special_k, all_peaks): # last number in each row is the total parts number of that person # the second last number in each row is the score of the overall configuration subset = -1 * np.ones((0, 20)) candidate = np.array([item for sublist in all_peaks for item in sublist]) for k in range(len(map_ids)): if k not in special_k: partAs = connection_all[k][:, 0] partBs = connection_all[k][:, 1] indexA, indexB = np.array(limb_seq[k]) - 1 for i in range(len(connection_all[k])): # = 1:size(temp,1) found = 0 subset_idx = [-1, -1] for j in range(len(subset)): # 1:size(subset,1): if subset[j][indexA] == partAs[i] or subset[j][indexB] == partBs[i]: subset_idx[found] = j found += 1 if found == 1: j = subset_idx[0] if (subset[j][indexB] != partBs[i]): subset[j][indexB] = partBs[i] subset[j][-1] += 1 subset[j][-2] += candidate[partBs[i].astype(int), 2] + connection_all[k][i][2] elif found == 2: # if found 2 and disjoint, merge them j1, j2 = subset_idx print("found = 2") membership = ((subset[j1] >= 0).astype(int) + (subset[j2] >= 0).astype(int))[:-2] if len(np.nonzero(membership == 2)[0]) == 0: # merge subset[j1][:-2] += (subset[j2][:-2] + 1) subset[j1][-2:] += subset[j2][-2:] subset[j1][-2] += connection_all[k][i][2] subset = np.delete(subset, j2, 0) else: # as like found == 1 subset[j1][indexB] = partBs[i] subset[j1][-1] += 1 subset[j1][-2] += candidate[partBs[i].astype(int), 2] + connection_all[k][i][2] # if find no partA in the subset, create a new subset elif not found and k < 17: row = -1 * np.ones(20) row[indexA] = partAs[i] row[indexB] = partBs[i] row[-1] = 2 row[-2] = sum(candidate[connection_all[k][i, :2].astype(int), 2]) + connection_all[k][i][2] subset = np.vstack([subset, row]) return subset, candidate def draw_key_point(subset, all_peaks, img_raw): del_ids = [] for i in range(len(subset)): if subset[i][-1] < 4 or subset[i][-2] / subset[i][-1] < 0.4: del_ids.append(i) subset = np.delete(subset, del_ids, axis=0) img_canvas = img_raw.copy() # B,G,R order for i in range(18): for j in range(len(all_peaks[i])): cv2.circle(img_canvas, all_peaks[i][j][0:2], 4, colors[i], thickness=-1) return subset, img_canvas def link_key_point(img_canvas, candidate, subset, stickwidth=4): for i in range(17): for n in range(len(subset)): index = subset[n][np.array(limb_seq[i]) - 1] if -1 in index: continue cur_canvas = img_canvas.copy() Y = candidate[index.astype(int), 0] X = candidate[index.astype(int), 1] mX = np.mean(X) mY = np.mean(Y) length = ((X[0] - X[1]) ** 2 + (Y[0] - Y[1]) ** 2) ** 0.5 angle = math.degrees(math.atan2(X[0] - X[1], Y[0] - Y[1])) polygon = cv2.ellipse2Poly((int(mY), int(mX)), (int(length / 2), stickwidth), int(angle), 0, 360, 1) cv2.fillConvexPoly(cur_canvas, polygon, colors[i]) img_canvas = cv2.addWeighted(img_canvas, 0.4, cur_canvas, 0.6, 0) return img_canvas def pad_right_down_corner(img, stride, pad_value): h = img.shape[0] w = img.shape[1] pad = 4 * [None] pad[0] = 0 # up pad[1] = 0 # left pad[2] = 0 if (h % stride == 0) else stride - (h % stride) # down pad[3] = 0 if (w % stride == 0) else stride - (w % stride) # right img_padded = img pad_up = np.tile(img_padded[0:1, :, :] * 0 + pad_value, (pad[0], 1, 1)) img_padded = np.concatenate((pad_up, img_padded), axis=0) pad_left = np.tile(img_padded[:, 0:1, :] * 0 + pad_value, (1, pad[1], 1)) img_padded = np.concatenate((pad_left, img_padded), axis=1) pad_down = np.tile(img_padded[-2:-1, :, :] * 0 + pad_value, (pad[2], 1, 1)) img_padded = np.concatenate((img_padded, pad_down), axis=0) pad_right = np.tile(img_padded[:, -2:-1, :] * 0 + pad_value, (1, pad[3], 1)) img_padded = np.concatenate((img_padded, pad_right), axis=1) return img_padded, pad if __name__ == '__main__': print(get_pose_model()) # First let's download the pre-trained model. # In[2]: # Using gdown to download the model directly from Google Drive #assert os.system(' conda install -y gdown') == 0 import gdown # In[3]: model = 'coco_pose_iter_440000.pth.tar' if not os.path.exists(model): url = 'https://drive.google.com/u/0/uc?export=download&confirm=f_Ix&id=0B1asvDK18cu_MmY1ZkpaOUhhRHM' gdown.download( url, model, quiet=False ) # In[4]: state_dict = torch.load(model)['state_dict'] # getting the pre-trained model's parameters # A state_dict is simply a Python dictionary object that maps each layer to its parameter tensor. model_pose = get_pose_model() # building the model (see fn. defn. above). To see the architecture, see below cell. model_pose.load_state_dict(state_dict) # Loading the parameters (weights, biases) into the model. model_pose.float() # I'm not sure why this is used. No difference if you remove it. if use_gpu is None: use_gpu = True if use_gpu: model_pose.cuda() model_pose = torch.nn.DataParallel(model_pose, device_ids=range(torch.cuda.device_count())) cudnn.benchmark = True def estimate_pose( img_ori, name=None, scale_param=None, display=None, ): if display is None: display = True if scale_param is None: scale_param = [0.5, 1.0, 1.5, 2.0] if display: if name is None: name = tempfile.mktemp( dir='/kaggle/working', suffix='.png', ) pprint.pprint( ['estimate_pose', dict(name=name)], ) # People might be at different scales in the image, perform inference at multiple scales to boost results # Predict Heatmaps for approximate joint position # Use Part Affinity Fields (PAF's) as guidance to link joints to form skeleton # PAF's are just unit vectors along the limb encoding the direction of the limb # A dot product of possible joint connection will be high if actual limb else low img_canvas = None img_points = None try: paf_info, heatmap_info = get_paf_and_heatmap(model_pose, img_ori, scale_param) peaks = extract_heatmap_info(heatmap_info) sp_k, con_all = extract_paf_info(img_ori, paf_info, peaks) subsets, candidates = get_subsets(con_all, sp_k, peaks) subsets, img_points = draw_key_point(subsets, peaks, img_ori) img_canvas = link_key_point(img_points, candidates, subsets) except ZeroDivisionError: pprint.pprint('zero de') img_points = img_ori.copy() img_canvas = img_ori.copy() # After predicting Heatmaps and PAF's, proceeed to link joints correctly if display: f = plt.figure(figsize=(15, 10)) plt.subplot(1, 2, 1) plt.imshow(img_points[...,::-1]) plt.subplot(1, 2, 2) plt.imshow(img_canvas[...,::-1]) f.savefig(name) return dict( img_points=img_points, img_canvas=img_canvas, ) # In[5]: return dict( cv2=cv2, estimate_pose=estimate_pose, model_pose=model_pose, ) def kernel_8( o_7, ): for i, o in enumerate([ '../input/indonesian-traditional-dance/tgagrakanyar/tga_00%d0.jpg' % k for k in range(6) ]): arch_image = o img_ori = o_7['cv2'].imread(arch_image) o_7['estimate_pose'](img_ori) def kernel_9_benchmark( o_7, ): import datetime t1 = o_7['cv2'].imread('../input/indonesian-traditional-dance/tgagrakanyar/tga_0000.jpg') t5 = 10 t2 = datetime.datetime.now() for k in range(t5): o_7['estimate_pose']( img_ori=t1, scale_param=[1.0], display=False, ) t3 = datetime.datetime.now() t4 = (t3 - t2).total_seconds() / t5 pprint.pprint( ['kernel_9_benchmark', dict(t4=t4, t5=t5)] ) def kernel_10(): import torch # Model model = torch.hub.load('ultralytics/yolov5', 'yolov5s') # or yolov5m, yolov5x, custom # Images img = 'https://ultralytics.com/images/zidane.jpg' # or file, PIL, OpenCV, numpy, multiple # Inference results = model(img) # Results results.print() # or .show(), .save(), .crop(), .pandas(), etc. return dict( model=model, ) def kernel_11_benchmark( o_7, o_10, ): import datetime t1 = o_7['cv2'].imread('../input/indonesian-traditional-dance/tgagrakanyar/tga_0000.jpg') t5 = 10 t2 = datetime.datetime.now() for k in range(t5): t6 = o_10['model'](t1) t7 = t6.pandas().xywhn t3 = datetime.datetime.now() t4 = (t3 - t2).total_seconds() / t5 pprint.pprint( ['kernel_11_benchmark', dict(t4=t4, t5=t5)] ) def kernel_13( o_6=None, ): t2 = [ '/kaggle/working', '/kaggle/input/garbage' ] t3 = [ os.path.join( o, 'kernel_13-object-detection.nc', ) for o in t2 ] t4 = [ o for o in t3 if os.path.exists(o) ] if not len(t4) > 0 or not o_6 is None: t1 = pandas.concat( sum( [ [ o2['t11'][0].assign( frame_id=k, video_path=o['video_path'] ) for k, o2 in enumerate(o['frames']) ] for o in o_6['t8'] ], [] ) ).to_xarray() t5 = t3[0] t1.to_netcdf(t5) del t1 elif len(t4) > 0: t5 = t4[0] else: raise NotImplementedError t1 = xarray.load_dataset(t5) return dict( t1=t1, ) def kernel_14( skip_o_6=None, skip_o_7_10_12_13=None, run_benchmark=None, ): if skip_o_6 is None: skip_o_6 = True if skip_o_7_10_12_13 is None: skip_o_7_10_12_13 = True if run_benchmark is None: run_benchmark = False o_3 = kernel_3(should_exist=True) o_4 = kernel_4(o_3=o_3) o_5 = kernel_5(o_4=o_4) if not skip_o_7_10_12_13: o_7 = kernel_7() o_10 = kernel_10() o_12 = kernel_12() else: o_7 = None o_10 = None o_12 = None if not skip_o_6: o_6 = kernel_6( o_7=None, o_10=o_10, o_12=None, max_frames=10000 ) else: o_6 = None if not skip_o_7_10_12_13: o_13 = kernel_13( o_6=o_6, ) else: o_13 = None if run_benchmark: o_11 = kernel_11_benchmark(o_7=o_7, o_10=o_10) o_9 = kernel_9_benchmark(o_7=o_7) o_8 = kernel_8(o_7=o_7) return dict( o_3=o_3, o_13=o_13, o_7=o_7, ) def kernel_15( o_14, ): t1 = pandas.DataFrame( numpy.unique( o_14['o_13']['t1']['name'].data, return_counts=True ) ).T pprint.pprint( dict( t1=t1, ) ) t29 = 'output-gif' if not os.path.exists(t29): os.makedirs(t29, exist_ok=True) for t2 in [ 'baseball glove', 'baseball bat', 'sports ball', 'person', ]: t28 = t2.replace(' ', '-') t3 = o_14['o_13']['t1'] t4 = numpy.where(t3.name.data == t2)[0] t30 = 'output-png/%s' % t28 if not os.path.exists(t30): os.makedirs(t30, exist_ok=True) numpy.random.seed(0) t22 = numpy.random.choice(t4, 10) pprint.pprint(t22) import tqdm t24 = [] t27 = [] for t5 in tqdm.tqdm(t22): t6 = t3.video_path.data[t5] t7 = t3.frame_id.data[t5] t8 = t3.to_dataframe().iloc[t5] #pprint.pprint([t6, t7]) #pprint.pprint(t8) import cv2 import matplotlib.pyplot t9 = cv2.VideoCapture(t6) t9.set(cv2.CAP_PROP_POS_FRAMES, t7) t10 = t9.read() t9.release() t11 = t10[1] t12 = cv2.cvtColor(t11, cv2.COLOR_BGR2RGB) t13 = t12.copy() t15 = numpy.array([t8.xcenter, t8.ycenter, t8.width, t8.height]) t16 = numpy.array([t13.shape[1], t13.shape[0], t13.shape[1], t13.shape[0]]) t17 = t15 * t16 t18 = t17[:2] - t17[2:] / 2 t19 = t17[:2] + t17[2:] / 2 t20 = numpy.array([ t18[0], t18[1], t19[0], t19[1], ]) t21 = numpy.round(t20).astype(numpy.int32) t14 = cv2.rectangle( t13, tuple(t21[:2]), tuple(t21[2:]), (0, 255, 0), 1, ) if False: t32 = o_14['o_7']['estimate_pose']( t12, scale_param=[1.0], display=False, )['img_canvas'] else: t32 = kernel_16([t12])['t6'][0] f = matplotlib.pyplot.figure(figsize=(8, 12)) f.suptitle( 'name %s, frame_id %d\nvideo_path %s' % ( t8['name'], t8.frame_id, t8.video_path, ) ) matplotlib.pyplot.subplot(2, 1, 1) matplotlib.pyplot.title( 'score %s' % ( t8.confidence, ) ) matplotlib.pyplot.imshow(t14) matplotlib.pyplot.subplot(2, 1, 2) matplotlib.pyplot.imshow(t32) t25 = os.path.join( t30, 'kernel_15-%s-%05d.jpg' % ( t28, t7, ) ) f.savefig(t25) t24.append(t25) matplotlib.pyplot.close(f) t27.append([t8, t21]) pprint.pprint( pandas.concat([ o[0] for o in t27 ], axis=1).T ) t23 = os.path.join( t29, 'output-%s.gif' % t28 ) if os.path.exists(t23): subprocess.check_call(['rm', t23]) subprocess.check_call( [ 'convert', '-delay', '100', '-loop', '0', *t24, t23, ] ) def kernel_16(images): assert isinstance(images, list) import cv2 import subprocess import os t2 = '/kaggle/working/test-input' subprocess.check_call([ 'rm', '-fr', t2, ]) subprocess.check_call([ 'mkdir', '-p', t2, ]) t1 = [] for i, o in enumerate(images): t5 = cv2.cvtColor(o, cv2.COLOR_RGB2BGR) t8 = 'image-%d.jpg' % i t3 = os.path.join( t2, t8, ) cv2.imwrite(t3, t5) t1.append( dict( image_name=t8, image_path=t3, image_canvas=o, image_index=i, ) ) t4 = '/kaggle/working/test-output' subprocess.check_call([ 'rm', '-fr', t4, ]) subprocess.check_call([ 'mkdir', '-p', t4, ]) with subprocess.Popen(''' cd /kaggle/working/AlphaPose && python3 \ scripts/demo_inference.py \ --cfg configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml \ --checkpoint pretrained_models/fast_res50_256x192.pth \ --indir %s \ --outdir %s \ --save_img ''' % (t2, t4), shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE) as p: pprint.pprint(p.communicate()) p.wait() assert p.returncode == 0 t6 = [] with io.open( os.path.join( t4, 'alphapose-results.json' ), 'r' ) as f: t7 = json.load(f) for o in t1: t9 = os.path.join( t4, 'vis', o['image_name'] ) assert os.path.exists(t9) t10 = cv2.imread(t9, cv2.IMREAD_COLOR) t11 = cv2.cvtColor(t10, cv2.COLOR_BGR2RGB) t6.append(t11) return dict( images=images, t1=t1, t6=t6, t7=t7, ) def kernel_17( o_14, max_images=None, ): if max_images is None: max_images = 10 t50 = [] for t2 in [ 'baseball glove', 'baseball bat', 'sports ball', 'person', ]: t28 = t2.replace(' ', '-') t3 = o_14['o_13']['t1'] t4 = numpy.where(t3.name.data == t2)[0] t30 = 'output-png/%s' % t28 if not os.path.exists(t30): os.makedirs(t30, exist_ok=True) numpy.random.seed(0) t22 = numpy.random.choice(t4, max_images) pprint.pprint(t22) import tqdm t24 = [] t27 = [] for t5 in tqdm.tqdm(t22): t6 = t3.video_path.data[t5] t7 = t3.frame_id.data[t5] t8 = t3.to_dataframe().iloc[t5] #pprint.pprint([t6, t7]) #pprint.pprint(t8) import cv2 import matplotlib.pyplot t9 = cv2.VideoCapture(t6) t9.set(cv2.CAP_PROP_POS_FRAMES, t7) t10 = t9.read() t9.release() t11 = t10[1] t12 = cv2.cvtColor(t11, cv2.COLOR_BGR2RGB) t13 = t12.copy() t50.append( dict( t2=t2, t28=t28, t6=t6, t5=t5, t7=t7, t12=t12, ) ) return dict( t50=t50, ) def kernel_18(o_17): t1 = [o['t12'] for o in o_17['t50']] t2 = kernel_16(t1) return dict( t2=t2, ) def kernel_19(o_18): kernel_25( o_18['t2']['t6'] ) def kernel_20( o_18, o_21=None, ): if o_21 is None: o_21 = kernel_21() import cv2 import numpy import os t1 = numpy.array(o_18['t2']['t7'][0]['keypoints']).reshape(17, -1) t2 = o_18['t2']['t6'][0] t3 = o_18['t2']['t1'][0]['image_canvas'].copy() assert o_18['t2']['t7'][0]['image_id'] == os.path.split(o_18['t2']['t1'][0]['image_name'])[1] for i, o2 in enumerate(o_21['p_color']): if i >= 17: print('fuck %d' % i) continue o1 = t1[i, :] cv2.circle(t3, tuple(o1[:2].astype(numpy.int32)), 3, o2, -1) cv2.imwrite('output.jpg', cv2.cvtColor(t3, cv2.COLOR_RGB2BGR)) cv2.imwrite('output-v2.jpg', cv2.cvtColor(t2, cv2.COLOR_RGB2BGR)) def kernel_21(): l_pair = [ (0, 1), (0, 2), (1, 3), (2, 4), # Head (5, 6), (5, 7), (7, 9), (6, 8), (8, 10), (17, 11), (17, 12), # Body (11, 13), (12, 14), (13, 15), (14, 16) ] p_color = [ # Nose, LEye, REye, LEar, REar (0, 255, 255), (0, 191, 255), (0, 255, 102), (0, 77, 255), (0, 255, 0), # LShoulder, RShoulder, LElbow, RElbow, LWrist, RWrist (77, 255, 255), (77, 255, 204), (77, 204, 255), (191, 255, 77), (77, 191, 255), (191, 255, 77), # LHip, RHip, LKnee, Rknee, LAnkle, RAnkle (204, 77, 255), (77, 255, 204), (191, 77, 255), (77, 255, 191), (127, 77, 255), (77, 255, 127), # Not Present ## Neck #(0, 255, 255), ] line_color = [(0, 215, 255), (0, 255, 204), (0, 134, 255), (0, 255, 50), (77, 255, 222), (77, 196, 255), (77, 135, 255), (191, 255, 77), (77, 255, 77), (77, 222, 255), (255, 156, 127), (0, 127, 255), (255, 127, 77), (0, 77, 255), (255, 77, 36)] #https://debuggercafe.com/wp-content/uploads/2020/10/keypoint_exmp.jpg # Seems like No Neck actually title = [ 'Nose', 'LEye', 'REye', 'LEar', 'REar', 'LShoulder', 'RShoulder', 'LElbow', 'RElbow', 'LWrist', 'RWrist', 'LHip', 'RHip', 'LKnee', 'RKnee', 'LAnkle', 'RAnkle', #'Neck', ] return dict( l_pair=l_pair, p_color=p_color, line_color=line_color, ) def kernel_22(o_18): t1 = o_18['t2']['t7'] t2 = [ numpy.array(o['keypoints']).reshape(17, 3) for o in t1 ] o_31 = kernel_31( image_id=[ o['image_id'] for o in t1 ], image_size=numpy.array([ [ list(o['image_canvas'].shape) for o in o_18['t2']['t1'] if o['image_name'] == t1[i]['image_id'] ][0] for i in range(len(t2)) ]), keypoints=numpy.stack(t2, axis=0), ) t12 = o_31['t12'] return dict( t4=t12, ) def kernel_24( img, keypoints, o_21=None, ): t3 = img.copy() if o_21 is None: o_21 = kernel_21() import cv2 import numpy import os for i, o2 in enumerate(o_21['p_color']): if i >= 17: print('fuck %d' % i) continue o1 = keypoints[i, :] cv2.circle(t3, tuple(o1[:2].astype(numpy.int32)), 3, o2, -1) return dict( t3=t3, ) def kernel_25(images, delay=None): if delay is None: delay = 100 import tqdm import os import cv2 import subprocess if len(images) == 0: pprint.pprint(['kernel_25', 'no images']) return t3 = 'kernel_25-output.dir' os.makedirs(t3, exist_ok=True) t6 = [] for i, o in tqdm.tqdm(enumerate(images)): t4 = 'image-%03d.jpg' % i t5 = os.path.join(t3, t4) t8 = cv2.cvtColor(o, cv2.COLOR_RGB2BGR) cv2.imwrite(t5, t8) t6.append(t5) t7 = os.path.join(t3, 'output.gif') if False: subprocess.check_call( [ 'convert', '-delay', '%d' % delay, '-loop', '0', *t6, t7, ] ) t8 = os.path.join( t3, 'output.mp4', ) if os.path.exists(t8): os.unlink(t8) if False: subprocess.check_call( [ 'ffmpeg', '-i', t7, t8, ] ) t9 = os.path.join(t3, 'input.txt') with io.open(t9, 'w') as f: f.write( '\n'.join( [ 'file %s' % os.path.split(o)[1] for o in t6 ] ) ) subprocess.check_call( [ 'ffmpeg', '-r', '%d' % int(100 / delay), '-f', 'concat', '-i', '%s' % t9, t8, ] ) def kernel_26(o_18, image_name): t1 = [ i for i, o in enumerate(o_18['t2']['t1']) if o['image_name'] == image_name ] assert len(t1) == 1 return t1[0] def kernel_23(o_18, o_22, ids=None): import numpy import tqdm if ids is None: ids = numpy.s_[:] t1 = numpy.arange(len(o_22['t4'])) t2 = t1[ids] pprint.pprint(t2[:5]) t7 = [] for o in tqdm.tqdm(t2): t3 = o_22['t4'][o] t9 = kernel_26(o_18=o_18, image_name=t3['image_name']) t4 = o_18['t2']['t1'][t9]['image_canvas'] t10 = o_18['t2']['t6'][t9] t4 = [ o['image_canvas'] for o in o_18['t2']['t1'] if o['image_name'] == t3['image_name'] ] assert len(t4) == 1 t5 = t4[0] t6 = kernel_24(t5, t3['keypoints']) t7.append(t6['t3']) t7.append(t10) kernel_25(t7) def kernel_27(): import tqdm import os import subprocess import pprint t5 = '/kaggle/working/ATL AT TOR - April 19, 2015-T0MUK91ZWys.mp4' t3 = '/kaggle/working/kernel_27-output.dir' os.makedirs(t3, exist_ok=True) #R = int(round(30 / 5)) FRAMERATE = 4 SLICE_LENGTH = 5 * 6 for i in tqdm.tqdm(range(100)): t1 = [SLICE_LENGTH * i, SLICE_LENGTH * (i + 1)] t2 = os.path.join(t3, 'slice-%d' % i) os.makedirs(t2, exist_ok=True) t4 = os.path.join(t2, 'output.mp4') if os.path.exists(t4): os.unlink(t4) with subprocess.Popen([ 'ffmpeg', '-i', '%s' % t5, '-filter:v', 'fps=fps=%d' % FRAMERATE, '-ss', '%d' % t1[0], '-t', '%d' % (t1[1] - t1[0]), '%s' % t4, ], stdout=subprocess.PIPE, stderr=subprocess.PIPE) as p: if False: pprint.pprint(p.communicate()) p.wait() assert p.returncode == 0 t6 = ''' cd /kaggle/working/AlphaPose && \ python3 \ scripts/demo_inference.py \ --cfg configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml \ --checkpoint pretrained_models/fast_res50_256x192.pth \ --video %s \ --outdir %s ''' % (t4, t2) if False: pprint.pprint([t4, t2, t6]) with subprocess.Popen(t6, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE) as p: if False: pprint.pprint(p.communicate()) p.wait() assert p.returncode == 0 def kernel_28( video_path=None, framerate=None, max_seconds=None, video_id=None, ): if video_id is None: video_id = '' import cv2 import tqdm import os import subprocess import pprint if framerate is None: framerate = 4 if max_seconds is None: max_seconds = 999999 if video_path is None: video_path = '/kaggle/working/ATL AT TOR - April 19, 2015-T0MUK91ZWys.mp4' t5 = video_path t3 = '/kaggle/working/kernel_28-output%s.dir' % video_id t13 = '/root/kernel_28-output.dir/tmp-slice' os.makedirs(t3, exist_ok=True) cap = None try: cap = cv2.VideoCapture(t5) fps = cap.get(cv2.CAP_PROP_FPS) # OpenCV2 version 2 used "CV_CAP_PROP_FPS" frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT)) real_duration = frame_count/fps duration = min(real_duration, max_seconds) pprint.pprint( dict( fps=fps, frame_count=frame_count, duration=duration, real_duration=real_duration, ) ) #R = int(round(30 / 5)) FRAMERATE = framerate SLICE_LENGTH = 5 * 6 for i in tqdm.tqdm(range(int(duration / SLICE_LENGTH + 1e-8))): t2 = os.path.join(t3, 'slice-%d' % i) if os.path.exists(t2): pprint.pprint(['t2', t2, 'exists', 'continue']) continue t1 = [SLICE_LENGTH * i, SLICE_LENGTH * (i + 1)] if os.path.exists(t13): subprocess.check_call(['rm', '-fr', t13]) os.makedirs(t13, exist_ok=True) t8 = numpy.array([t1[0] * fps, t1[1] * fps]).astype(numpy.int32) cap.set(cv2.CAP_PROP_POS_FRAMES, t8[0]) t12 = numpy.arange(t8[0], t8[1], SLICE_LENGTH * FRAMERATE) for k in range(t8[1] - t8[0]): ret, frame = cap.read() t11 = cap.get(cv2.CAP_PROP_POS_FRAMES) if numpy.isin(t11, t12): t10 = os.path.join( t13, 'frame-%d.jpg' % t11, ) cv2.imwrite(t10, frame) os.makedirs(t2, exist_ok=True) t6 = ''' cd /kaggle/working/AlphaPose && \ python3 \ scripts/demo_inference.py \ --cfg configs/coco/resnet/256x192_res50_lr1e-3_1x.yaml \ --checkpoint pretrained_models/fast_res50_256x192.pth \ --indir %s \ --outdir %s ''' % (t13, t2) if False: pprint.pprint([t4, t2, t6]) with subprocess.Popen(t6, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE) as p: if False: pprint.pprint(p.communicate()) p.wait() assert p.returncode == 0 finally: if not cap is None: cap.release() def kernel_29(video_path=None, video_id=None,): if video_id is None: video_id = '' if video_path is None: video_path = '/kaggle/working/ATL AT TOR - April 19, 2015-T0MUK91ZWys.mp4' assert os.path.exists(video_path) t40 = 'kernel_29-poses%s.json' % video_id t4 = os.path.join('/kaggle/working', t40) t6 = [ t4, os.path.join('/kaggle/input/garbage', t40), ] t7 = [ o for o in t6 if os.path.exists(o) ] if len(t7) == 0: t1 = [ dict( data=json.load( io.open( o, 'r' ) ), input_path=o ) for o in glob.glob( '/kaggle/working/kernel_28-output%s.dir/slice-*/*.json' % video_id ) ] assert len(t1) > 0 with io.open(t4, 'w') as f: f.write(json.dumps(t1)) t7.append(t4) with io.open(t7[0], 'r') as f: t1 = json.load(f) t8 = sum([ o['data'] for o in t1 ], []) t10 = re.compile('frame-(\d+)\.jpg') for i, o in enumerate(t8): o['frame_id'] = int(t10.match(o['image_id'])[1]) t9 = sorted( t8, key=lambda o: o['frame_id'] ) t2 = pandas.DataFrame(t9) t5 = t2.to_xarray() t5['keypoints'] = xarray.DataArray( [ numpy.array(o).reshape(17, 3) for o in t5['keypoints'].data ], dims=['index', 'joint', 'feature'], ) o_31 = kernel_31( image_id=t5.image_id.data, image_size=[ kernel_32(video_path) ] * t5.index.shape[0], keypoints=t5.keypoints.data, ) for k, v in o_31['t13'].items(): #pprint.pprint([k, v]) t5[k] = v return dict( t5=t5, t2=t2, t9=t9, ) def kernel_30( o_29, ids=None, delay=None, prev_frames=None, max_frames=None, max_ids=None, video_path=None, low_portion=None, low_mean_conf=None, no_dots=None, ): if no_dots is None: no_dots = False if low_portion is None: low_portion = 0.1 if low_mean_conf is None: low_mean_conf = 0.6 if video_path is None: video_path = '/kaggle/working/ATL AT TOR - April 19, 2015-T0MUK91ZWys.mp4' if max_frames is None: max_frames = 9999 if max_ids is None: max_ids = 70 if prev_frames is None: prev_frames = 0 t5 = o_29['t5'] if delay is None: delay = 200 if ids is None: numpy.random.seed(0) ids = numpy.random.choice( t5.index.shape[0], 10 ) elif ids == 'v1': t8 = numpy.where( numpy.logical_and( o_29['t5'].portion > low_portion, o_29['t5'].mean_conf > low_mean_conf, ) )[0] ids = numpy.random.choice( t8, min(len(t8), 70), replace=False, ) elif ids == 'v2': t8 = numpy.stack( [ o_29['t5'].portion > 0.1, o_29['t5'].mean_conf > 0.6, o_29['t5']['t11'].data, ], axis=0 ).prod(axis=0) pprint.pprint([t8.sum(), t8.mean()]) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), ) elif ids == 'v3': t8 = numpy.stack( [ o_29['t5'].portion > 0.1, o_29['t5'].mean_conf > 0.6, o_29['t5']['t10'].data, ], axis=0 ).prod(axis=0) pprint.pprint([t8.sum(), t8.mean()]) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), ) elif ids == 'v4': t8 = numpy.stack( [ o_29['t5'].portion > 0.1, o_29['t5'].min_conf > 0.6, o_29['t5']['t9'].data > 0, ], axis=0 ).prod(axis=0) pprint.pprint([t8.sum(), t8.mean()]) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), ) elif ids == 'v5': t8 = numpy.stack( [ o_29['t5'].portion > 0.1, o_29['t5'].min_conf > 0.5, o_29['t5']['t8'].data > 0, ], axis=0 ).prod(axis=0) pprint.pprint([t8.sum(), t8.mean()]) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), ) elif ids == 'v6': t8 = numpy.stack( [ o_29['t5'].portion > 0.02, o_29['t5'].min_conf > 0.4, o_29['t5']['t8'].data > 0, ], axis=0 ).prod(axis=0) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), replace=False, ) pprint.pprint([ t8.sum(), t8.mean(), ids, o_29['t5'].sel(index=o_29['t5'].index.data[ids[:5]]).to_dict(), ]) elif ids == 'v7': t8 = numpy.stack( [ o_29['t5'].portion > 0.02, o_29['t5'].min_conf > 0.4, numpy.stack( [ o_29['t5']['t4'].data > 0, o_29['t5']['t5'].data > 0, ], axis=0 ).sum(axis=0) > 0, ], axis=0 ).prod(axis=0) ids = numpy.random.choice( numpy.where(t8)[0], min(70, t8.sum()), replace=False, ) pprint.pprint([ t8.sum(), t8.mean(), ids, o_29['t5'].sel(index=o_29['t5'].index.data[ids[:5]]).to_dict(), ]) else: assert isinstance(ids, numpy.ndarray) ids = ids[:max_ids] pprint.pprint(['ids', ids]) t7 = [] t16 = video_path t15 = kernel_32(t16) for o in tqdm.tqdm(ids): t2 = t5.keypoints.data[o] t3 = t5.frame_id.data[o] cap = None try: cap = cv2.VideoCapture(t16) if prev_frames > 0: t14 = kernel_24( numpy.zeros(t15, dtype=numpy.uint8), t2, )['t3'] t17 = (t14 == 0).astype(numpy.uint8) for k in range(prev_frames): cap.set(cv2.CAP_PROP_POS_FRAMES, t3 - prev_frames + k) ret, frame = cap.read() t12 = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) #t13 = kernel_24(t12, t2)['t3'] t13 = t12 * t17 + t14 t7.append(t13) cap.set(cv2.CAP_PROP_POS_FRAMES, t3) ret, frame = cap.read() t4 = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB) if not no_dots: t6 = kernel_24(t4, t2)['t3'] else: t6 = t4 t7.append(t6) finally: if not cap is None: cap.release() kernel_25(t7[:max_frames], delay=delay) def kernel_31(image_id, image_size, keypoints): def get_angle(a,b): from math import sqrt, acos, degrees, atan, degrees #print(a) #print(b) del_y = a[1]-b[1] del_x = b[0]-a[0] if del_x == 0: del_x = 0.1 #print("Del_X : "+str(del_x)+"-----Del_Y: "+str(del_y)) angle = 0 if del_x > 0 and del_y > 0: angle = degrees(atan(del_y / del_x)) elif del_x < 0 and del_y > 0: angle = degrees(atan(del_y / del_x)) + 180 return angle def angle_gor(a,b,c,d): import numpy as np from math import sqrt, acos, degrees, atan, degrees ab=[a[0]-b[0],a[1]-b[1]] ab1=[c[0]-d[0],c[1]-d[1]] cos = \ abs( ab[0] * ab1[0] + ab[1] * ab1[1] ) / ( sqrt( ab[0] ** 2 + ab[1] ** 2 ) * \ sqrt( ab1[0] ** 2 + ab1[1] ** 2 ) + 1e-8 ) ang = acos(cos) return ang*180/np.pi def sit_ang(a,b,c,d): ang=angle_gor(a,b,c,d) s1=0 if ang != None: #print("Angle",ang) if ang < 120 and ang>40: s1=1 return s1 def sit_rec(a,b,c,d): from math import sqrt, acos, degrees, atan, degrees ab = [a[0] - b[0], a[1] - b[1]] ab1 = [c[0] - d[0], c[1] - d[1]] l1=sqrt(ab[0]**2+ab[1]**2) l2=sqrt(ab1[0]**2+ab1[1]**2) s=0 if l1!=0 and l2!=0: #print(l1,l2, "---------->>>") if l2/l1>=1.5: s=1 return s t3 = [] for i, o in enumerate(keypoints): t4 = numpy.min(o[:, 0]) t5 = numpy.max(o[:, 0]) t6 = numpy.min(o[:, 1]) t7 = numpy.max(o[:, 1]) t8 = (t5 - t4) * (t7 - t6) t9 = image_size[i] t10 = t9[0] * t9[1] t3.append( dict( t8=t8, t9=t9, t10=t10, ) ) t12 = [ dict( square=t3[i]['t8'], total=t3[i]['t10'], portion=t3[i]['t8'] / (t3[i]['t10'] + 1e-8), keypoints=keypoints[i], image_name=image_id[i], ) for i in range(len(t3)) ] for i, o in enumerate(t12): t1 = o['keypoints'] t2 = get_angle( t1[3, :2], t1[4, :2], ) t3 = get_angle( t1[6, :2], t1[7, :2], ) t4 = 0 if 30 < t2 and t2 < 150: t4 = 1 t5 = 0 if 30 < t3 and t3 < 150: t5 = 1 t6 = t4 + t6 t7 = 0 if t6 == 1: t7 = 1 t8 = 0 t8 += sit_rec( t1[9, :2], t1[10, :2], t1[10, :2], t1[11, :2], ) t8 += sit_rec( t1[12, :2], t1[13, :2], t1[13, :2], t1[14, :2], ) t9 = 0 t9 += sit_ang( t1[9, :2], t1[10, :2], t1[10, :2], t1[11, :2], ) t9 += sit_ang( t1[12, :2], t1[13, :2], t1[13, :2], t1[14, :2], ) t10 = 0 if t8 > 0 or t9 > 0: t10 = 1 t11 = 0 if t8 == 0 and t9 == 0: t11 = 1 o.update( dict( t2=t2, t3=t3, t4=t4, t5=t5, t6=t6, t7=t7, t8=t8, t9=t9, t10=t10, t11=t11, mean_conf=t1[:, 2].mean(), min_conf=t1[:, 2].min(), ) ) t13 = pandas.DataFrame(t12).drop(columns=['keypoints']).to_xarray() return dict( t12=t12, t13=t13, ) def kernel_32(video_path): assert os.path.exists(video_path) cap = None try: cap = cv2.VideoCapture(video_path) ret, frame = cap.read() return frame.shape finally: if not cap is None: cap.release() def kernel_33(): o_14 = kernel_14() o_15 = kernel_15(o_14=o_14) assert os.system(r''' python repo/d1/mlb_player_v5.py ''') == 0 o_17 = kernel_17(o_14=o_14, max_images=100) o_18 = kernel_18(o_17=o_17) o_19 = kernel_19(o_18=o_18) o_20 = kernel_20(o_18=o_18) o_22 = kernel_22(o_18=o_18) import pandas o_23 = kernel_23( o_18=o_18, o_22=o_22, ids=pandas.DataFrame( o_22['t4'] ).query('portion > 0.1').index.values ) o_27 = kernel_27() o_28 = kernel_28() o_29 = kernel_29() import numpy o_30 = kernel_30(o_29=o_29, ids='v1') def kernel_34(o_14): t1 = numpy.unique( o_14['o_3']['t5']['events']['event'], return_counts=True, ) t2 = [ [o, o2] for o, o2 in zip(*t1) ] pprint.pprint( t2 ) def kernel_35(): return r''' 1. https://posetrack.net/leaderboard.php 2. https://posetrack.net/users/view_entry_details.php?entry=CF_chl2 3. https://posetrack.net/users/view_entry_details.php?entry=DCPose_chl2 4. https://github.com/Pose-Group/DCPose/blob/main/docs/Installation.md 5. https://github.com/openpifpaf/openpifpaf_posetrack 6. https://github.com/Guanghan/lighttrack 7. https://github.com/leoxiaobin/deep-high-resolution-net.pytorch 8. https://github.com/CMU-Perceptual-Computing-Lab/openpose https://github.com/CMU-Perceptual-Computing-Lab/openpose/issues/1736#issuecomment-736846794 ''' def kernel_36(): # -*- coding: utf-8 -*- """OpenPose.ipynb Automatically generated by Colaboratory. Original file is located at https://colab.research.google.com/drive/1VDjRLKAu9KLQky0gv4RLJjeH6NRPiWXy """ assert os.system(r''' cd openpose && rm -rf build || true && mkdir build && cd build && cmake .. && make -j`nproc` ''') == 0 """# Pose Detection with OpenPose This notebook uses an open source project [CMU-Perceptual-Computing-Lab/openpose](https://github.com/CMU-Perceptual-Computing-Lab/openpose.git) to detect/track multi person poses on a video from your Google Drive @dinatih update base on https://colab.research.google.com/github/tugstugi/dl-colab-notebooks/blob/master/notebooks/OpenPose.ipynb ## Choose a video from your Google Drive """ from google.colab import drive drive.mount('/content/drive') """## Install OpenPose on Google Colab """ import os from os.path import exists, join, basename, splitext git_repo_url = 'https://github.com/CMU-Perceptual-Computing-Lab/openpose.git' project_name = splitext(basename(git_repo_url))[0] if 1 or not exists(project_name): assert os.system(r'''!rm -rf openpose''') == 0 # see: https://github.com/CMU-Perceptual-Computing-Lab/openpose/issues/949 print("install new CMake becaue of CUDA10") cmake_version = 'cmake-3.20.2-linux-x86_64.tar.gz' if not exists(cmake_version): assert os.system(r'''!wget -q 'https://cmake.org/files/v3.20/{cmake_version}' ''') == 0 assert os.system(r'''!tar xfz {cmake_version} --strip-components=1 -C /usr/local ''') == 0 print("clone openpose") assert os.system(r'''!git clone -q --depth 1 $git_repo_url ''') == 0 print("install system dependencies") assert os.system(r'''!apt-get -qq install -y libatlas-base-dev libprotobuf-dev libleveldb-dev libsnappy-dev libhdf5-serial-dev protobuf-compiler libgflags-dev libgoogle-glog-dev liblmdb-dev opencl-headers ocl-icd-opencl-dev libviennacl-dev ''') == 0 print("build openpose") assert os.system(r'''!cd openpose && rm -rf build || true && mkdir build && cd build && cmake .. && make -j`nproc` ''') == 0 """## From a Google Drive's folder""" import os from os.path import exists, join, basename, splitext folder_path = '/content/drive/My Drive/openpose/' files = os.listdir(folder_path) files.reverse() for filename in files: if filename.endswith('.mp4') and not filename.endswith('-openpose.mp4'): print(filename) colab_video_path = folder_path + filename print(colab_video_path) colab_openpose_video_path = colab_video_path.replace('.mp4', '') + '-openpose.mp4' print(colab_openpose_video_path) if not exists(colab_openpose_video_path): assert os.system(r'''!cd openpose && ./build/examples/openpose/openpose.bin --hand --face --number_people_max 12 --video '{colab_video_path}' --display 0 --write_video_with_audio --write_video '{colab_openpose_video_path}' # --net_resolution "-1x736" --scale_number 4 --scale_gap 0.25 ''') == 0 """## From Youtube (Downloaded to your Drive)""" assert os.system(r'''!pip install youtube-dl ''') == 0 youtube_id = '2021-05-07_22-00-55_UTC' assert os.system(r'''!youtube-dl -f mp4 -o '/content/drive/My Drive/openpose/%(id)s.mp4' {youtube_id} ''') == 0 colab_video_path = '/content/drive/My Drive/openpose/' + youtube_id + '.mp4' colab_openpose_video_path = colab_video_path.replace('.mp4', '') + '-openpose.mp4' assert os.system(r'''!cd openpose && ./build/examples/openpose/openpose.bin --number_people_max 12 --video '{colab_video_path}' --display 0 --write_video_with_audio --write_video '{colab_openpose_video_path}' # --net_resolution "-1x736" --scale_number 4 --scale_gap 0.25 ''') == 0 """## Run OpenPose""" assert os.system(r'''!wget 'https://cdn-eu-hz-4.ufile.io/get/z5yes0he?token=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' -O output.mp4 ''') == 0 # video_folder = os.path.dirname(colab_video_path) # video_base_name = os.path.basename(colab_video_path) # print(video_base_name) import os from os.path import exists, join, basename, splitext #colab_video_path = '/content/drive/My Drive/bachata.mp4' colab_video_path = '/content/output.mp4' colab_openpose_video_path = colab_video_path.replace('.mp4', '') + '-openpose.mp4' assert os.system(r'''!cd openpose && ./build/examples/openpose/openpose.bin --number_people_max 12 --video '{colab_video_path}' --display 0 --write_video_with_audio --write_video '{colab_openpose_video_path}' # --net_resolution "-1x736" --scale_number 4 --scale_gap 0.25 ''') == 0 assert os.system(r'''!cmake -version ''') == 0 def kernel_37(): return r''' 1. https://www.youtube.com/channel/UCcTy8yBARva3WzrmIcaaWPA/about 2. https://www.play-cricket.com/website/results/4721802 ''' def kernel_38( video_path, framerate=None, max_seconds=None, ids=None, prev_frames=None, delay=None, low_mean_conf=None, low_portion=None, no_dots=None, ): if ids is None: ids = 'v7' if prev_frames is None: prev_frames = 59 if delay is None: delay = 3 t2 = hashlib.sha256( video_path.encode('utf-8') ).hexdigest() t1 = '/kaggle/working/video%s.mp4' % t2 if not os.path.exists(t1): subprocess.check_call( [ 'youtube-dl', '-f', '18', video_path, '-o', t1, ] ) kernel_28( t1, framerate=framerate, max_seconds=max_seconds, video_id=t2, ) o_29 = kernel_29( t1, video_id=t2, ) o_30 = kernel_30( o_29=o_29, ids=ids, delay=delay, prev_frames=prev_frames, max_frames=9999, max_ids=999, video_path=t1, low_mean_conf=low_mean_conf, low_portion=low_portion, no_dots=no_dots, ) return dict( o_29=o_29, o_30=o_30, )