TorchSigΒΆ
TorchSig is an open-source signals processing machine learning toolkit.
- Datasets
- Transforms
- Transforms
- Augmentations
- DatasetBasebandMixUp
- DatasetBasebandCutMix
- CutOut
- PatchShuffle
- DatasetWidebandMixUp
- DatasetWidebandCutMix
- SpectrogramRandomResizeCrop
- RandomResample
- RandomTimeShift
- TimeCrop
- TimeReversal
- AmplitudeReversal
- RandomFrequencyShift
- RandomDelayedFrequencyShift
- LocalOscillatorDrift
- GainDrift
- AutomaticGainControl
- IQImbalance
- RollOff
- AddSlope
- SpectralInversion
- ChannelSwap
- RandomMagRescale
- RandomDropSamples
- Quantize
- Clip
- RandomConvolve
- TargetSNR
- AddNoise
- TimeVaryingNoise
- RayleighFadingChannel
- ImpulseInterferer
- RandomPhaseShift
- SpectrogramDropSamples
- SpectrogramPatchShuffle
- SpectrogramTranslation
- SpectrogramMosaicCrop
- SpectrogramMosaicDownsample
- Target Transforms
- DescToClassName
- DescToClassNameSNR
- DescToClassIndex
- DescToClassIndexSNR
- DescToMask
- DescToMaskSignal
- DescToMaskFamily
- DescToMaskClass
- DescToSemanticClass
- DescToBBox
- DescToAnchorBoxes
- DescPassThrough
- DescToBinary
- DescToCustom
- DescToClassEncoding
- DescToWeightedMixUp
- DescToWeightedCutMix
- DescToBBoxDict
- DescToBBoxSignalDict
- DescToBBoxFamilyDict
- DescToInstMaskDict
- DescToSignalInstMaskDict
- DescToSignalFamilyInstMaskDict
- DescToListTuple
- ListTupleToDesc
- LabelSmoothing
- Models
- Signal Data Utilities