Repository navigation
Conversation
…on. AudioGenerator is updated to interact with this abstraction instead of directly with the Onnx session(s). Also, AutoEncoder specifies both an encode and decode operation, which will eventually allow for wider use throughout the platform. Minor improvements to OnnxDitModel and OnnxFeatures.
…oder. Minor improvements to OnnxAutoEncoder and AudioGenerator. Introduced WaveData::loadMultiChannel and improved the efficiency of WavFile::channel.
… set up with different strategies for generating feature data. DefaultWaveDataFeatureProvider is introduced to capture the previous approach using WaveData::features, while AutoEncoderFeatureProvider is introduced to compute the features using AutoEncoder::encode. Introduced AutoEncoder::getSampleRate, AutoEncoder::getLatentSampleRate and WaveData::getDefaultFeatureSettings.
…lection will continue to return the data for a single channel for backwards compatibility, but the new method WaveData::getData will return data for all channels. Eliminated WaveData::allocateCollection and AudioLibrary::load methods. Introduced OnnxFeatures::defaultOptions and a new constructor for OnnxAutoEncoder.
…ata if existing WaveDetails are provided and updated AudioLibrary::getDetails so that it computes feature data if it is missing. Removed unnecessary warning from WaveData constructor.
… input to stereo. Improved error handling in OnnxFeatures.
…WaveDataProvider that accepts a File.
…ed to combine the feature data of multiple samples using a linear sum, where the scale for each feature component is determined by the dot product of a weight vector with an input vector. AutoEncoder::encode and AutoEncoder::decode are updated to operate on PackedCollection Producer instead of PackedCollection.
…hat weights are generated and used correctly and that feature shapes are consistent. AudioModulator is updated to use AudioComposer. OnnxAutoEncoder::decode is updated to handle omitting the batch dimension.
…re now uniform across the entire space of features, and ComposableAudioFeatures::getResultant cuts off scale values below zero.
…s and updated WaveDetailsFactory::similarity to attempt to use feature data instead of frequency data, if it is available.
…bed dimensions and seed properties for AudioModel. Introduced AutoEncoder::getMaximumDuration.
…peration. Reduced logging in WaveData and marked WaveData::getCollection deprecated. Test updates.
…m and introduced ConditionalAudioScoring. New methods of OnnxFeatures.
…lAudioScoring. Introduced DitModel::getAttentionActivations and updated DiffusionTransformer to implement it. Improved error handling in OnnxFeatures.
…o in ConditionalAudioScoring.
…lsFactory::forWaveData to tolerate null WaveData.
…of WaveDetails for processing via a dedicated ExecutorService.
…alueSequenceCell, SineWaveCell and related types. Eliminated ValueSequenceData. Updated AudioFilterData::output to ensure that the correct shape is returned. Adjusted configuration of AudioSceneOptimizer.
…n AudioScene, eliminating the need for ProjectedGenomeSet. Accordingly, AutomationManager, ChordProgressionManager, DefaultChannelSectionFactory, EfxManager, MixdownManager, PatternLayerManager, RiseManager and SceneSectionManager are all updated to no longer require a dedicated ProjectedGenome. OptimizeFactorFeatures methods are updated to return a List of ProjectedGenes instead of expecting a dedicated ProjectedChromosome. Many minor test updates.
…and PatternLayerManager.
…ate and feature (output) sample rate. Introduced WaveDataProvider::getChannelCount and WaveDataProvider::get supporting resampling. Updated TreeNoteSource::computeProviders to avoid excluding sources based on sample rate. WaveDetailsFactory now uses dot product for feature similarity calculation. Introduced warning about identical features in AudioLibrary::computeSimilarities.
…for pause and resume of lower priority detail computation jobs.
… moving the resulting file(s) into place when the serialization process is complete. Uses of LibraryDestination are updated to ensure that this final flush operation is performed when the process is over.
…new PackedCollection is returned each time.
…addition Computation. Cleanup in AudioSceneOptimizer.
Release Candidate for Version 0.41
…nough data to compute similarity.
…rting progress for WaveDetailJobs. Updated WaveDetailsFactory to align the orientation of difference similarity and product similarity, with both measures using positive values to represent greater similarity. Introduced WaveDetails::getValidFeatureFrameCount. Some test updates.
…f outlier values. Logging adjustments.
…DetailsJob::toString.
…n AnnotatedAudioRendererTest.
…se operation with PassThroughProducer and introduced related test cases. Updated AudioModulator to use the same seed so that the results are reproducible for validating changes.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.