# DiffusePast: Diffusion-based Generative Replay for Class Incremental Semantic Segmentation

arXiv: https://arxiv.org/abs/2308.01127 · HTML: https://mlsanity.com/paper/2308.01127/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models.
- Builds on 41 of its 50 references that are in the map.
- First posted on arXiv 2 Aug 2023.

## Placed in

- Diffusion models / [Medical imaging and segmentation](https://mlsanity.com/diffusion/medical-imaging-and-segmentation/)

## Builds on (most cited first, 30 of 41)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Denoising Diffusion Implicit Models](https://mlsanity.com/paper/2010.02502/) (2020)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding](https://mlsanity.com/paper/2205.11487/) (2022)
- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [Adding Conditional Control to Text-to-Image Diffusion Models](https://mlsanity.com/paper/2302.05543/) (2023)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Large Scale GAN Training for High Fidelity Natural Image Synthesis](https://mlsanity.com/paper/1809.11096/) (2018)
- [Learning without Forgetting](https://mlsanity.com/paper/1606.09282/) (2016)
- [An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion](https://mlsanity.com/paper/2208.01618/) (2022)
- [SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations](https://mlsanity.com/paper/2108.01073/) (2021)
- [Continual Learning with Deep Generative Replay](https://mlsanity.com/paper/1705.08690/) (2017)
- [Label-Efficient Semantic Segmentation with Diffusion Models](https://mlsanity.com/paper/2112.03126/) (2021)
- [Synthetic Data from Diffusion Models Improves ImageNet Classification](https://mlsanity.com/paper/2304.08466/) (2023)
- [Is synthetic data from generative models ready for image recognition?](https://mlsanity.com/paper/2210.07574/) (2022)
- [Improving Robustness using Generated Data](https://mlsanity.com/paper/2110.09468/) (2021)
- [Class-Incremental Learning: A Survey](https://mlsanity.com/paper/2302.03648/) (2023)
- [Modeling the Background for Incremental Learning in Semantic Segmentation](https://mlsanity.com/paper/2002.00718/) (2020)
- [PLOP: Learning without Forgetting for Continual Semantic Segmentation](https://mlsanity.com/paper/2011.11390/) (2020)
- [DiffuMask: Synthesizing Images with Pixel-level Annotations for Semantic Segmentation Using Diffusion Models](https://mlsanity.com/paper/2303.11681/) (2023)
- [Incremental Learning Techniques for Semantic Segmentation](https://mlsanity.com/paper/1907.13372/) (2019)
- [Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA](https://mlsanity.com/paper/2304.06027/) (2023)
- [Semantic Image Synthesis via Diffusion Models](https://mlsanity.com/paper/2207.00050/) (2022)
- [Learning Semantic Segmentation from Synthetic Data: A Geometrically Guided Input-Output Adaptation Approach](https://mlsanity.com/paper/1812.05040/) (2018)
- [BigDatasetGAN: Synthesizing ImageNet with Pixel-wise Annotations](https://mlsanity.com/paper/2201.04684/) (2022)
- [Less-forgetting Learning in Deep Neural Networks](https://mlsanity.com/paper/1607.00122/) (2016)
- [Continual Semantic Segmentation via Repulsion-Attraction of Sparse and Disentangled Latent Representations](https://mlsanity.com/paper/2103.06342/) (2021)

## Built on by (most cited first, 8 of 8)

- [Continual Forgetting for Pre-trained Vision Models](https://mlsanity.com/paper/2403.11530/) (2024)
- [A Survey on Continual Semantic Segmentation: Theory, Challenge, Method and Application](https://mlsanity.com/paper/2310.14277/) (2023)
- [Future-Proofing Class-Incremental Learning](https://mlsanity.com/paper/2404.03200/) (2024)
- [Practical Continual Forgetting for Pre-trained Vision Models](https://mlsanity.com/paper/2501.09705/) (2025)
- [Avoid Catastrophic Forgetting with Rank-1 Fisher from Diffusion Models](https://mlsanity.com/paper/2509.23593/) (2025)
- [CLoRA: Parameter-Efficient Continual Learning with Low-Rank Adaptation](https://mlsanity.com/paper/2507.19887/) (2025)
- [Trust Region Continual Learning as an Implicit Meta-Learner](https://mlsanity.com/paper/2602.02417/) (2026)
- [Stabilizing language models under continual learning via condition-anchored distillation](https://mlsanity.com/paper/2610.06940/) (2026)

## Abstract (arXiv)

The Class Incremental Semantic Segmentation (CISS) extends the traditional segmentation task by incrementally learning newly added classes. Previous work has introduced generative replay, which involves replaying old class samples generated from a pre-trained GAN, to address the issues of catastrophic forgetting and privacy concerns. However, the generated images lack semantic precision and exhibit out-of-distribution characteristics, resulting in inaccurate masks that further degrade the segmentation performance. To tackle these challenges, we propose DiffusePast, a novel framework featuring a diffusion-based generative replay module that generates semantically accurate images with more reliable masks guided by different instructions (e.g., text prompts or edge maps). Specifically, DiffusePast introduces a dual-generator paradigm, which focuses on generating old class images that align with the distribution of downstream datasets while preserving the structure and layout of the original images, enabling more precise masks. To adapt to the novel visual concepts of newly added classes continuously, we incorporate class-wise token embedding when updating the dual-generator. Moreover, we assign adequate pseudo-labels of old classes to the background pixels in the new step images, further mitigating the forgetting of previously learned knowledge. Through comprehensive experiments, our method demonstrates competitive performance across mainstream benchmarks, striking a better balance between the performance of old and novel classes.