root_path = '/var/tmp/stoehr/Template-Speech-Recognition/' data_path = root_path + 'Data/' exp_path = root_path + 'Notebook/1/' tmp_data_path = exp_path + 'data/' import sys, os, cPickle sys.path.append(root_path) import template_speech_rec.get_train_data as gtrd import template_speech_rec.estimate_template as et train_data_path = root_path+'Data/Train/' file_indices = gtrd.get_data_files_indices(train_data_path) syllable = np.array(['aa','r']) avg_bgd, syllable_examples, backgrounds = gtrd.get_syllable_examples_backgrounds_files(train_data_path, file_indices, syllable, num_examples=-1, verbose=True) clipped_bgd = np.clip(avg_bgd.E,.1,.4) np.save(tmp_data_path+'clipped_bgd_102012.npy',clipped_bgd) padded_examples, lengths = et.extend_examples_to_max(clipped_bgd,syllable_examples, return_lengths=True) aar_template, aar_registered = et.register_templates_time_zero(syllable_examples,min_prob=.01) test_example_lengths = gtrd.get_detect_lengths(data_path+'Test/') np.save(tmp_data_path+'test_example_lengths_102012.npy',test_example_lengths)
num_parts = 100 old_devel_data_path = root_path + 'Development/092412/data/' parts = np.load(old_devel_data_path+'bm_templates%d_%d.npy' % (lower_cutoff,num_parts)) log_part_blocks, log_invpart_blocks = gtrd.reorg_parts_for_fast_filtering(parts) log_part_blocks = log_part_blocks.astype(np.float32) log_invpart_blocks = log_invpart_blocks.astype(np.float32) file_indices = gtrd.get_data_files_indices(train_data_path) syllable = np.array(['b','ay']) avg_bgd, syllable_examples, backgrounds = gtrd.get_syllable_examples_backgrounds_files(train_data_path, file_indices, syllable, num_examples=-1, verbose=True) parts_avg_bgd, parts_syllable_examples, parts_backgrounds = gtrd.get_syllable_examples_backgrounds_files(train_data_path, file_indices, syllable, log_part_blocks, log_invpart_blocks, num_examples=-1, verbose=True)