# Making Training-Free Diffusion Segmentors Scale with the Generative Power

arXiv: https://arxiv.org/abs/2603.06178 · HTML: https://mlsanity.com/paper/2603.06178/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Medical imaging and segmentation, one of 14 directions in Diffusion models.
- Builds on 39 of its 53 references that are in the map.
- First posted on arXiv 6 Mar 2026.

## Placed in

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

## Builds on (most cited first, 30 of 39)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Diffusion Models Beat GANs on Image Synthesis](https://mlsanity.com/paper/2105.05233/) (2021)
- [The Cityscapes Dataset for Semantic Urban Scene Understanding](https://mlsanity.com/paper/1604.01685/) (2016)
- [SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis](https://mlsanity.com/paper/2307.01952/) (2023)
- [CLIPScore: A Reference-free Evaluation Metric for Image Captioning](https://mlsanity.com/paper/2104.08718/) (2021)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [Extract Free Dense Labels from CLIP](https://mlsanity.com/paper/2112.01071/) (2021)
- [Label-Efficient Semantic Segmentation with Diffusion Models](https://mlsanity.com/paper/2112.03126/) (2021)
- [Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models](https://mlsanity.com/paper/2303.04803/) (2023)
- [PixArt-Sigma: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation](https://mlsanity.com/paper/2403.04692/) (2024)
- [A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Correspondence](https://mlsanity.com/paper/2305.15347/) (2023)
- [Gated Attention for Large Language Models: Non-linearity, Sparsity, and Attention-Sink-Free](https://mlsanity.com/paper/2505.06708/) (2025)
- [Unleashing Text-to-Image Diffusion Models for Visual Perception](https://mlsanity.com/paper/2303.02153/) (2023)
- [What the DAAM: Interpreting Stable Diffusion Using Cross Attention](https://mlsanity.com/paper/2210.04885/) (2022)
- [Diffusion Hyperfeatures: Searching Through Time and Space for Semantic Correspondence](https://mlsanity.com/paper/2305.14334/) (2023)
- [Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing](https://mlsanity.com/paper/2403.03431/) (2024)
- [ReCo: Retrieve and Co-segment for Zero-shot Transfer](https://mlsanity.com/paper/2206.07045/) (2022)
- [Diffusion Model as Representation Learner](https://mlsanity.com/paper/2308.10916/) (2023)
- [Diffusion Model is Secretly a Training-free Open Vocabulary Semantic Segmenter](https://mlsanity.com/paper/2309.02773/) (2023)
- [Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2404.05384/) (2024)
- [Diffusion 3D Features (Diff3F): Decorating Untextured Shapes with Distilled Semantic Features](https://mlsanity.com/paper/2311.17024/) (2023)
- [ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features](https://mlsanity.com/paper/2502.04320/) (2025)
- [MoH: Multi-Head Attention as Mixture-of-Head Attention](https://mlsanity.com/paper/2410.11842/) (2024)
- [R&B: Region and Boundary Aware Zero-shot Grounded Text-to-image Generation](https://mlsanity.com/paper/2310.08872/) (2023)
- [DiffCut: Catalyzing Zero-Shot Semantic Segmentation with Diffusion Features and Recursive Normalized Cut](https://mlsanity.com/paper/2406.02842/) (2024)
- [Not All Diffusion Model Activations Have Been Evaluated as Discriminative Features](https://mlsanity.com/paper/2410.03558/) (2024)
- [A Unified Perspective for Loss-Oriented Imbalanced Learning via Localization](https://mlsanity.com/paper/2310.04752/) (2023)

## Built on by (most cited first, 3 of 3)

- [Generative Semantic Segmentation via an Observable Semantic-Image Interface and Hierarchical Generator Evidence Alignment](https://mlsanity.com/paper/2608.11537/) (2026)
- [Diffuse2Seg: Diffusion Models Can Segment Anything Without Supervision](https://mlsanity.com/paper/2609.06491/) (2026)
- [We Have Been Interpreting Diffusion Transformer Backbone Behaviors Wrong](https://mlsanity.com/paper/2610.14517/) (2026)

## Abstract (arXiv)

As powerful generative models, text-to-image diffusion models have recently been explored for discriminative tasks. A line of research focuses on adapting a pre-trained diffusion model to semantic segmentation without any further training, leading to training-free diffusion segmentors. These methods typically rely on cross-attention maps from the model's attention layers, which are assumed to capture semantic relationships between image pixels and text tokens. Ideally, such approaches should benefit from more powerful diffusion models, i.e., stronger generative capability should lead to better segmentation. However, we observe that existing methods often fail to scale accordingly. To understand this issue, we identify two underlying gaps: (i) cross-attention is computed across multiple heads and layers, but there exists a discrepancy between these individual attention maps and a unified global representation. (ii) Even when a global map is available, it does not directly translate to accurate semantic correlation for segmentation, due to score imbalances among different text tokens. To bridge these gaps, we propose two techniques: auto aggregation and per-pixel rescaling, which together enable training-free segmentation to better leverage generative capability. We evaluate our approach on standard semantic segmentation benchmarks and further integrate it into a generative technique, demonstrating both improved performance broad applicability. Codes are at https://github.com/Darkbblue/goca.