# recog_sq.append(label)
# target_sq.append(target)
recog_sq[str(k)].append(label)
target_sq[str(k)].append(target)
k += 1
start_To = start_To + S_enc_frame
for m in range(len(recog_sq.keys())):
target = target_sq[str(m)][1:]
recog = recog_sq[str(m)][1:]
f_name = vid.split('/')[-1].split('.')[0]
path = args.results_save_path + "/obs" + str(int(
args.S_enc)) + "-pred" + str(m)
write_predictions(path, f_name, recog)
path = args.results_save_path + "/obs" + str(int(
args.S_enc)) + "-pred" + str(m) + "_target"
write_predictions(path, f_name + "_target", target)
# report = classification_report(target, recog,output_dict=True)
# print ("Accuracy: " + str(round(accuracy_score(target,recog) * 100,2)))
# print ("Precision " + str(round(report['macro avg']['precision']*100,2)))
# print ("Recall " + str(round(report['macro avg']['recall']*100,2)))
# print ("f1-score " + str( round(report['macro avg']['f1-score']*100,2)))
# n_T=np.zeros(len(actions_dict.keys()))
# n_F=np.zeros(len(actions_dict.keys()))
# for i in range(len(target)):
# if target[i]==recog[i]:
# n_T[actions_dict[target[i]]]+=1 # Se la classe per questo frame e stata riconosciuta aggiugno 1 a True
# else:
# n_F[actions_dict[target[i]]]+=1 # Se la calss di questo frame e errata aggiungo uno a False
for pred_p in pred_percentages:
pred_len = int(pred_p*vid_len)
output_len = pred_len + len(observed_content)
label_seq, length_seq = get_label_length_seq(observed_content)
with tf.Session() as sess:
label_seq, length_seq = model.predict(sess, model_restore_path, pred_len, label_seq, length_seq, actions_dict, T)
recognition = []
for i in range(len(label_seq)):
recognition = np.concatenate((recognition, [label_seq[i]]*int(length_seq[i])))
recognition = recognition[:output_len]
#write results to file
f_name = vid.split('/')[-1].split('.')[0]
path=args.results_save_path+"/obs"+str(obs_p)+"-pred"+str(pred_p)
write_predictions(path, f_name, recognition)
elif args.model == "cnn":
model = ModelCNN(args.nRows, nClasses)
for vid in list_of_videos:
f_name = vid.split('/')[-1].split('.')[0]
observed_content=[]
vid_len = 0
if args.input_type == "gt":
file_ptr = open(vid, 'r')
content = file_ptr.read().split('\n')[:-1]
vid_len = len(content)
for obs_p in obs_percentages:
if args.input_type == "decoded":