def use_hmm(img_times, change_vals, fps=10, min_secs_for_train_to_pass=8):
from sklearn.hmm import GaussianHMM
X = np.column_stack(change_vals)
n_components = 2
model = GaussianHMM(n_components, covariance_type="diag", n_iter=1000)
model.fit([X.T])
#thresh = 10**-15
#model.transmat_ = np.array([[1-thresh,thresh],[1-thresh,thresh]])
hidden_states = model.predict(X.T)
# print trained parameters and plot
print("Transition matrix")
print(model.transmat_)
print()
print("means and vars of each hidden state")
for i in range(n_components):
print("%dth hidden state" % i)
print("mean = ", model.means_[i])
print("var = ", np.diag(model.covars_[i]))
print()
if model.means_[0][0] > model.means_[1][0]: # assume most most frames have no train, switch labels if necessary
hidden_states = 1 - hidden_states
train_spotted = filter_out_short_motions(hidden_states, min_secs_for_train_to_pass, fps)
plot_timeline(img_times, change_vals, hidden_states, train_spotted)
utils.copy_image_subset(config.experiment_data_frames, config.experiment_output_frames_hmm, np.nonzero(train_spotted)[0])
return train_spotted
def save_positive_negative_examples(img_times, train_spotted, n=20):
train_inds, = np.nonzero(train_spotted)
non_train_inds, = np.nonzero(1-train_spotted)
positive_inds = train_inds[::len(train_inds)/(2*n)]
negative_inds = non_train_inds[::len(non_train_inds)/(2*n)]
utils.copy_image_subset(config.experiment_data_frames, os.path.join(config.experiment_output, 'positive_examples'), positive_inds[2::+1])
utils.copy_image_subset(config.experiment_data_frames, os.path.join(config.experiment_output, 'positive_examples'), positive_inds[2::])
utils.copy_image_subset(config.experiment_data_frames, os.path.join(config.experiment_output, 'negative_examples'), negative_inds)
