def generate_precision_recall_text(mode='val'):
model = load_moviescope_model('text')
yTrue, plotFeatures = gather_features(mode, return_video=False, reverse=True)
plotFeatures = np.array(map(list, zip(*plotFeatures)))
yPreds = model.predict([plotFeatures[0], plotFeatures[1]])
dump_pkl((yTrue, yPreds), mode+'_pred_text')
return
def generate_precision_recall_video(mode='val'):
model = load_moviescope_model('wiki_im_video_sgd')
yTrue, videoFeatures = gather_features(mode, return_plot=False)
_, videoFeatures = augment_labels_lstm(yTrue, videoFeatures, 200)
yPreds = model.predict(videoFeatures)
dump_pkl((yTrue, yPreds), mode+'_pred_video_sgd')
return
def generate_precision_recall_vislang(mode='val', merge_mode='sum'):
if merge_mode == 'bilinear':
model = vislang_model(merge_mode)
model.load_weights('data/weights/weights_min_loss_%s.h5' % merge_mode)
else:
model = load_moviescope_model('eq_VisLang_%s' % merge_mode)
yTrue, plotFeatures, videoFeatures = gather_features(mode, reverse=True)
plotFeatures = np.array(map(list, zip(*plotFeatures)))
_, videoFeatures = augment_labels_lstm(yTrue, videoFeatures, 200)
yPreds = model.predict([videoFeatures, plotFeatures[0], plotFeatures[1]])
dump_pkl((yTrue, yPreds), mode+'_pred_eq_vislang_'+merge_mode)
def return_confident_results(mode='val'):
model = load_moviescope_model('wiki_im_VisLang')
genrePredictionDict = dict((i,[]) for i in range(26))
textObj = load_pkl('plot_object_train')
labels, plotFeatures, videoFeatures, movieIds = gather_features(mode, return_id=True)
_, videoFeatures = augment_labels_lstm(labels, videoFeatures, 200)
predictionScores = model.predict([videoFeatures, plotFeatures])
for index in range(len(predictionScores)):
for i in range(26):
genrePredictionDict[i].append((predictionScores[index][i],movieIds[index]))
dump_pkl(genrePredictionDict, 'genrePredictionDict_'+mode)
for i in range(26):
print sorted(genrePredictionDict[i], reverse=True)[:10]
return
prediction_scores = model.predict(videoFeatures,verbose=0)
prediction_scores = np.mean(prediction_scores, axis=0)
predCount = np.bincount(predictions)
return prediction_scores, predCount
def test_video(videoPath):
videoFeatures = np.array(list(extract_feature_video(videoPath, verbose=True)))
print trial_video(videoFeatures)
def test_videos(testPath):
videoPaths = glob(testPath+'/*')
for videoPath in videoPaths:
test_video(videoPath)
def trials(genre):
genreFeatures = gather_videos(genre,limit_videos=40)
for videoFeatures in genreFeatures:
count = trial_video(videoFeatures)
print count
if __name__=="__main__":
model = load_moviescope_model('spatial_3g_bs32_ep50_nf_75')
# model = load_moviescope_model('all_bs_16_ep_50_nf_35')
from time import time
s = time()
test_video(argv[1])
print time()-s,"seconds."
"%\n")
print("Recall of " + genres[i] + " is " + str(round(rec, 2)) + "%\n")
print("F1 of " + genres[i] + " is " + str(round(f1, 2)) + "%\n")
print("Accuracy of " + genres[i] + " is " + str(round(acc, 2)) + "%\n")
print("---------------")
total_acc = total_acc + acc
total_acc = total_acc / 3
print("Overall Accuracy is " + str(round(total_acc, 2)) + "%\n")
#
if __name__ == "__main__":
from sys import argv
if (argv[1] == 'test'):
model = load_moviescope_model(argv[2])
extractor = argv[3]
ultimate_evaluate(model, extractor)
elif (argv[1] == 'video'):
"""to call test_video"""
genres, scores = test_video(argv[2], argv[3])
#print("genresArr", genres)
print("jumlah per genre: ", np.bincount(genres))
predictedGenre = np.argmax(np.bincount(genres))
genreDict = {0: 'action', 1: 'horror', 2: 'romance'}
frameSequence = ' | '.join([genreDict[key] for key in genres])
print(predictedGenre)
print(frameSequence)
print("\nGenre of this video is ", genreDict[predictedGenre])
def test_vislang(mode='val'):
model = load_moviescope_model('wiki_im_vislang')
labels, plotFeatures, videoFeatures = gather_features(mode)
evaluate_vislang(model, videoFeatures, plotFeatures, labels)
def test(mode='val'):
valLabels, plotFeatures, videoFeatures = gather_features(mode)
plotModel = load_moviescope_model(best_plot_model)
videoModel = load_moviescope_model(best_video_model)
evaluate_text(plotModel, plotFeatures, valLabels)
evaluate_visual(videoModel, videoFeatures, valLabels)
def test_text():
valLabels, valFeatures = gather_features(mode='val', return_video=False)
model = load_moviescope_model(best_plot_model)
evaluate_text(model, valFeatures, valLabels)
def test_visual():
valLabels, valFeatures = gather_features('val', return_plot=False)
model = load_moviescope_model(best_video_model)
evaluate_visual(model, valFeatures, valLabels)