# Joint Time-Frequency Scattering

arXiv: https://arxiv.org/abs/1807.08869 · HTML: https://mlsanity.com/paper/1807.08869/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Jul 2018.

## Built on by (most cited first, 11 of 11)

- [Scattering Networks on the Sphere for Scalable and Rotationally Equivariant Spherical CNNs](https://mlsanity.com/paper/2102.02828/) (2021)
- [Parametric Scattering Networks](https://mlsanity.com/paper/2107.09539/) (2021)
- [Leveraging Pre-Trained Autoencoders for Interpretable Prototype Learning of Music Audio](https://mlsanity.com/paper/2402.09318/) (2024)
- [Deep scattering network for speech emotion recognition](https://mlsanity.com/paper/2105.04806/) (2021)
- [Mesostructures: Beyond Spectrogram Loss in Differentiable Time-Frequency Analysis](https://mlsanity.com/paper/2301.10183/) (2023)
- [A Methodology for Exploring Deep Convolutional Features in Relation to Hand-Crafted Features with an Application to Music Audio Modeling](https://mlsanity.com/paper/2106.00110/) (2021)
- [Differentiable Time-Frequency Scattering on GPU](https://mlsanity.com/paper/2204.08269/) (2022)
- [Perceptual-Neural-Physical Sound Matching](https://mlsanity.com/paper/2301.02886/) (2023)
- [Mean-Field Microcanonical Gradient Descent](https://mlsanity.com/paper/2403.08362/) (2024)
- [SigWavNet: Learning Multiresolution Signal Wavelet Network for Speech Emotion Recognition](https://mlsanity.com/paper/2502.00310/) (2025)
- [SCRAPL: Scattering Transform with Random Paths for Machine Learning](https://mlsanity.com/paper/2602.11145/) (2026)

## Abstract (arXiv)

In time series classification and regression, signals are typically mapped into some intermediate representation used for constructing models. Since the underlying task is often insensitive to time shifts, these representations are required to be time-shift invariant. We introduce the joint time-frequency scattering transform, a time-shift invariant representation which characterizes the multiscale energy distribution of a signal in time and frequency. It is computed through wavelet convolutions and modulus non-linearities and may therefore be implemented as a deep convolutional neural network whose filters are not learned but calculated from wavelets. We consider the progression from mel-spectrograms to time scattering and joint time-frequency scattering transforms, illustrating the relationship between increased discriminability and refinements of convolutional network architectures. The suitability of the joint time-frequency scattering transform for time-shift invariant characterization of time series is demonstrated through applications to chirp signals and audio synthesis experiments. The proposed transform also obtains state-of-the-art results on several audio classification tasks, outperforming time scattering transforms and achieving accuracies comparable to those of fully learned networks.