Initialize parameters for the evaluator. Parameters ---------- See StackedNoteEvaluator...

Initialize parameters for the evaluator. Parameters ---------- See StackedEvaluator class for others... pit

Initialize parameters for the evaluator. Parameters ---------- See StackedPitchListEvaluator class...

Initialize parameters for the evaluator. Parameters ---------- See Evaluator class for others... profile :

Initialize the common properties among instrument profiles. Parameters ---------- low : int Lowest midi no

Initialize the common properties among tablature profiles. Parameters ---------- tuning : list of str Name

Initialize the profile and establish parameter defaults in function signature. Parameters ---------- tuning : list o

Initialize parameters for the waveform visualizer. Parameters ---------- See Visualizer class for others...

Initialize parameters for the TFR visualizer. Parameters ---------- See Visualizer class for others... sampl

Initialize parameters for the stacked pitch list visualizer. Parameters ---------- See Visualizer class for others..

Initialize parameters for the guitar tablature visualizer. Parameters ---------- See Visualizer class for others...

Initialize parameters for the pianoroll visualizer. Parameters ---------- See TFRVisualizer class for others...

Initialize parameters for the HVQT. Parameters ---------- See FeatureModule & VQT class for others... fmin :

Initialize parameters for the waveform wrapper. Parameters ---------- See FeatureModule class for others...

Initialize parameters for computing signal power. Parameters ---------- See WaveformWrapper class...

Initialize parameters for the HCQT. Parameters ---------- See HVQT class...

Initialize parameters for the streaming wrapper. Parameters ---------- module : FeatureModule Feature extr

Initialize parameters for the microphone streaming interface. Parameters ---------- See FeatureStream class for othe

Initialize parameters for the audio file streaming interface. Parameters ---------- See AudioStream class for others

Initialize parameters for the Mel Spectrogram. Parameters ---------- See WaveformWrapper class for others...

Initialize parameters for the Mel Spectrogram. Parameters ---------- See STFT class for others... n_mels : i

Initialize parameters common to all feature extraction modules. Parameters ---------- sample_rate : int or float

Initialize parameters for the feature combination. Parameters ---------- modules : list of FeatureModules

Initialize parameters for the VQT. Parameters ---------- See FeatureModule class for others... fmin : float

Initialize the dataset and establish parameter defaults in function signature. Parameters ---------- See Transcripti

Call the parent constructor.

Call the parent constructor.

Call the parent constructor.

Initialize parameters common to all datasets as fields and instantiate as a PyTorch Dataset. Parameters ----------

Initialize the dataset and establish parameter defaults in function signature. Parameters ---------- See Transcripti

Initialize the model and establish parameter defaults in function signature. Parameters ---------- See Transcription

Initialize parameters common to all models and instantiate model as a PyTorch Module. Parameters ----------

Initialize fields of the multi-label softmax layer. Parameters ---------- dim_in : int Dimensionality of i

Initialize fields of the multi-label logistic layer. Parameters ---------- dim_in : int Dimensionality of

Initialize the model and establish parameter defaults in function signature. Parameters ---------- See OnsetsFrames

Initialize the acoustic model and establish parameter defaults in function signature. Parameters ---------- dim_in :

Initialize the language model and establish parameter defaults in function signature. Parameters ---------- dim_in :

Initialize the language model and establish parameter defaults in function signature. Parameters ---------- See Lang

Defines the notion of length for the dataset - used by PyTorch Dataset class. Returns ---------- length : int

Obtain a list of possible splits. Currently, the splits are by player, but we could equally do genre, key, etc., as long as get_track

Obtain a list of possible splits. Currently, the splits are train/validation/test. Returns ---------- splits : list

Return the average of the currently tracked results across all evaluators. Returns ---------- average : dictionary

Convert batched notes from MIDI to Hertz. Parameters ---------- batched_notes : ndarray (N x 3) Array of note intervals and MI

Convert batched notes from Hertz to MIDI. Parameters ---------- batched_notes : ndarray (N x 3) Array of note intervals and He

Convert a collection of (batched) notes into a dictionary of stacked notes. Parameters ---------- batched_notes : ndarray (N x 3)

Blur activations by convolving them with a kernel. Parameters ---------- activations : ndarray Provided activations kernel

Concatenate two collections of stacked notes. Parameters ---------- stacked_notes : dict Dictionary containing (slice -> (pitc

Concatenate two stacked pitch lists, appending observations with new times and blending observations with preexisting times. This functio

Change the top-level directory from the path chain of each file. Parameters ---------- new_dir : string New top-level director

Obtain the time elapsed since the given system time. Parameters ---------- start_time : float Arbitrary system time decima

Determine if a pitch list representation contains empty observations. Parameters ---------- pitch_list : list of ndarray (N x [...])

Determine if a set of intervals contains any overlap. Parameters ---------- intervals : ndarray (N x 2) Array of onset-offset

Determine if a pitch list representation contains overlapping pitch contours. Parameters ---------- pitch_list : list of ndarray (N

Append together matching entries of two dictionaries. This will deliberately skip tuples, as in stacked representations. TODO - may be re

Detach gradient computation for all tensors. Parameters ---------- track : dict Dictionary containing data for a track Re

Collapse unnecessary dimensions of an array or tensor. Parameters ---------- track : dict Dictionary containing data for a tra

Convert all tensor entries in a dictionary to ndarray. Parameters ---------- track : dict Dictionary containing data for a tra

Add all tensor entries in a dictionary to a specified device. Parameters ---------- track : dict Dictionary containing data fo

Convert all ndarray entries in a dictionary to a specified type. Parameters ---------- track : dict Dictionary containing data

Convert all ndarray entries in a dictionary to tensors. Parameters ---------- track : dict Dictionary containing data for a tr

Add a new dimension to an array or tensor. Parameters ---------- track : dict Dictionary containing data for a track dim :

Download GuitarSet to a specified location. Parameters ---------- save_dir : string Directory in which to

Download MAESTRO version 1 to a specified location. Parameters ---------- save_dir : string Directory in w

Download MAESTRO version 2 to a specified location. Parameters ---------- save_dir : string Directory in w

Download MAESTRO version 3 to a specified location. Parameters ---------- save_dir : string Directory in w

Currently, this function stops execution. I am not aware of a way to automatically download MAPS. I will consider this again some tim

Each constituent dataset will be downloaded, if necessary, at the time of it's initialization. Parameters ----------

Here, just simply pass through the raw multi pitch data. Parameters ---------- raw_output : dict Dictionar

Estimate notes for each slice of a stacked multi pitch activation map. Parameters ---------- raw_output : dict

Track notes for each slice of a stacked multi pitch activation map, acquiring note estimates when the notes are complete. TO

Estimate notes from a multi pitch activation map. Parameters ---------- raw_output : dict Dictionary conta

Track notes within a multi pitch activation map, acquiring note estimates when the notes are complete. TODO - verify transcr

Refine a stacked multi pitch activation map. Parameters ---------- raw_output : dict Dictionary containing

Refine a multi pitch activation map. Parameters ---------- raw_output : dict Dictionary containing raw out

Convert stacked multi pitch activations to stacked pitch lists. Parameters ---------- raw_output : dict Di

Convert a multi pitch activation map to a pitch list. Parameters ---------- raw_output : dict Dictionary c

Convert tablature into a stacked multi pitch activation map. Parameters ---------- raw_output : dict Dicti

Convert a stacked multi pitch activation map into a single multi pitch representation. Parameters ---------- raw_out

Convert stacked notes into a single notes representation. Parameters ---------- raw_output : dict Dictiona

Convert a stacked pitch list into a single representation. Parameters ---------- raw_output : dict Diction

Convert a stacked multi pitch activation map into a stacked onsets activation map. Parameters ---------- raw_output

Convert a stacked multi pitch activation map into a stacked offsets activation map. Parameters ---------- raw_output

This should not be called directly on a ComboEvaluator.

Simply return loss in a new results dictionary. Parameters ---------- estimated : ndarray Single loss valu

Evaluate a stacked multi pitch estimate with respect to a reference. Parameters ---------- estimated : ndarray (S x

Evaluate a multi pitch estimate with respect to a reference. Parameters ---------- estimated : ndarray (F x T)

Evaluate stacked note estimates with respect to a reference. Parameters ---------- estimated : dict Dictio

Evaluate note estimates with respect to a reference. Parameters ---------- estimated : ndarray (N x 3) Arr

Evaluate stacked pitch list estimates with respect to a reference. Parameters ---------- estimated : dict

Evaluate pitch list estimates with respect to a reference. Parameters ---------- estimated : tuple containing

Evaluate a stacked multi pitch tablature estimate with respect to a reference. Parameters ---------- estimated : nda

Evaluate class membership estimates with respect to a reference. Parameters ---------- estimated : ndarray (S x T)

Acquire the next frame of features from the stream. Returns ---------- features : ndarray Features for one

Augment file names for sorting since, e.g., /'500/' will by default be scored as higher than /'1500/'. One way to fix this is by adding the

Remove any note duplicates, where a duplicate is defined as an entry with the same pitch and onset time. If there are duplicates, we keep

Remove any note duplicates within each slice, where a duplicate is defined as an entry with the same pitch and onset time. Parameters