# MDSGen: Fast and Efficient Masked Diffusion Temporal-Aware Transformers for Open-Domain Sound Generation

arXiv: https://arxiv.org/abs/2410.02130 · HTML: https://mlsanity.com/paper/2410.02130/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Audio, speech and music, one of 14 directions in Diffusion models.
- Builds on 47 of its 55 references that are in the map.
- First posted on arXiv 3 Oct 2024.

## Placed in

- Diffusion models / [Audio, speech and music](https://mlsanity.com/diffusion/audio-speech-and-music/)

## Builds on (most cited first, 30 of 47)

- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2304.08818/) (2023)
- [SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition](https://mlsanity.com/paper/1904.08779/) (2019)
- [ImageBind: One Embedding Space To Bind Them All](https://mlsanity.com/paper/2305.05665/) (2023)
- [HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis](https://mlsanity.com/paper/2010.05646/) (2020)
- [MaskGIT: Masked Generative Image Transformer](https://mlsanity.com/paper/2202.04200/) (2022)
- [DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Models](https://mlsanity.com/paper/2211.01095/) (2022)
- [Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video Generators](https://mlsanity.com/paper/2303.13439/) (2023)
- [VGGSound: A Large-scale Audio-Visual Dataset](https://mlsanity.com/paper/2004.14368/) (2020)
- [AudioLDM: Text-to-Audio Generation with Latent Diffusion Models](https://mlsanity.com/paper/2301.12503/) (2023)
- [Masked Autoencoders that Listen](https://mlsanity.com/paper/2207.06405/) (2022)
- [MDTv2: Masked Diffusion Transformer is a Strong Image Synthesizer](https://mlsanity.com/paper/2303.14389/) (2023)
- [SoundNet: Learning Sound Representations from Unlabeled Video](https://mlsanity.com/paper/1610.09001/) (2016)
- [Adan: Adaptive Nesterov Momentum Algorithm for Faster Optimizing Deep Models](https://mlsanity.com/paper/2208.06677/) (2022)
- [Taming Visually Guided Sound Generation](https://mlsanity.com/paper/2110.08791/) (2021)
- [Diff-Foley: Synchronized Video-to-Audio Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2306.17203/) (2023)
- [SoundStorm: Efficient Parallel Audio Generation](https://mlsanity.com/paper/2305.09636/) (2023)
- [Seeing and Hearing: Open-domain Visual-Audio Generation with Diffusion Latent Aligners](https://mlsanity.com/paper/2402.17723/) (2024)
- [FoleyCrafter: Bring Silent Videos to Life with Lifelike and Synchronized Sounds](https://mlsanity.com/paper/2407.01494/) (2024)
- [I Hear Your True Colors: Image Guided Audio Generation](https://mlsanity.com/paper/2211.03089/) (2022)
- [Localizing Visual Sounds the Easy Way](https://mlsanity.com/paper/2203.09324/) (2022)
- [Free-Bloom: Zero-Shot Text-to-Video Generator with LLM Director and LDM Animator](https://mlsanity.com/paper/2309.14494/) (2023)
- [Generating Visually Aligned Sound from Videos](https://mlsanity.com/paper/2008.00820/) (2020)
- [Cascade RPN: Delving into High-Quality Region Proposal Network with Adaptive Convolution](https://mlsanity.com/paper/1909.06720/) (2019)
- [VampNet: Music Generation via Masked Acoustic Token Modeling](https://mlsanity.com/paper/2307.04686/) (2023)
- [A3T: Alignment-Aware Acoustic and Text Pretraining for Speech Synthesis and Editing](https://mlsanity.com/paper/2203.09690/) (2022)

## Built on by (most cited first, 9 of 9)

- [TARO: Timestep-Adaptive Representation Alignment with Onset-Aware Conditioning for Synchronized Video-to-Audio Synthesis](https://mlsanity.com/paper/2504.05684/) (2025)
- [Policy Learning from Large Vision-Language Model Feedback without Reward Modeling](https://mlsanity.com/paper/2507.23391/) (2025)
- [Occlusion-robust Stylization for Drawing-based 3D Animation](https://mlsanity.com/paper/2508.00398/) (2025)
- [E-MD3C: Taming Masked Diffusion Transformers for Efficient Zero-Shot Object Customization](https://mlsanity.com/paper/2502.09164/) (2025)
- [PAVAS: Physics-Aware Video-to-Audio Synthesis](https://mlsanity.com/paper/2512.08282/) (2025)
- [A Hidden Semantic Bottleneck in Conditional Embeddings of Diffusion Transformers](https://mlsanity.com/paper/2602.21596/) (2026)
- [AirSplat: Alignment and Rating for Robust Feed-Forward 3D Gaussian Splatting](https://mlsanity.com/paper/2603.25129/) (2026)
- [SALSA-V: Shortcut-Augmented Long-form Synchronized Audio from Videos](https://mlsanity.com/paper/2510.02916/) (2025)
- [Model-Guided Dual-Role Alignment for High-Fidelity Open-Domain Video-to-Audio Generation](https://mlsanity.com/paper/2510.24103/) (2025)

## Abstract (arXiv)

We introduce MDSGen, a novel framework for vision-guided open-domain sound generation optimized for model parameter size, memory consumption, and inference speed. This framework incorporates two key innovations: (1) a redundant video feature removal module that filters out unnecessary visual information, and (2) a temporal-aware masking strategy that leverages temporal context for enhanced audio generation accuracy. In contrast to existing resource-heavy Unet-based models, textttMDSGen employs denoising masked diffusion transformers, facilitating efficient generation without reliance on pre-trained diffusion models. Evaluated on the benchmark VGGSound dataset, our smallest model (5M parameters) achieves 97.9% alignment accuracy, using 172× fewer parameters, 371% less memory, and offering 36× faster inference than the current 860M-parameter state-of-the-art model (93.9% accuracy). The larger model (131M parameters) reaches nearly 99% accuracy while requiring 6.5× fewer parameters. These results highlight the scalability and effectiveness of our approach. The code is available at https://bit.ly/mdsgen.