File size: 3,260 Bytes
0862dd7
 
 
 
 
3568b5a
0862dd7
3568b5a
 
0862dd7
3568b5a
 
0862dd7
3568b5a
 
 
0862dd7
3568b5a
0862dd7
 
 
 
 
3568b5a
0862dd7
3568b5a
 
 
 
 
 
0862dd7
 
 
 
3568b5a
0862dd7
 
3568b5a
0862dd7
 
 
 
 
 
 
 
 
 
 
3568b5a
 
 
 
0862dd7
 
 
3568b5a
0862dd7
 
 
 
 
 
3568b5a
0862dd7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3568b5a
 
0862dd7
 
 
 
 
3568b5a
 
 
 
 
 
 
0862dd7
 
 
 
 
6d326f0
0862dd7
 
6d326f0
 
0862dd7
 
 
 
 
 
 
 
 
 
 
 
3568b5a
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
---
tags:
- text-to-image
- diffusers
widget:
- text: a photo of a laptop above a dog
  output:
    url: images/laptop-above-dog.jpg
- text: a photo of a potted plant to the right of a motorcycle
  output:
    url: images/potted_plant-right-motorcycle.jpg
- text: a photo of a sheep below a sink
  output:
    url: images/sheep-below-sink.jpg
base_model: runwayml/stable-diffusion-v1-5
license: apache-2.0
---
# CoMPaSS-SD1.5

<Gallery />

## Model description 

# CoMPaSS-SD1.5

\[[Project Page]\]
\[[code]\]
\[[arXiv]\]

A UNet that enhances spatial understanding capabilities of the StableDiffusion 1.5 text-to-image
diffusion model.  This model demonstrates significant improvements in generating images with specific
spatial relationships between objects.

## Model Details

- **Base Model**: StableDiffusion 1.5
- **Training Data**: SCOP dataset (curated from COCO)
- **Framework**: Diffusers
- **License**: Apache-2.0 (see [./LICENSE])

## Intended Use

- Generating images with accurate spatial relationships between objects
- Creating compositions that require specific spatial arrangements
- Enhancing the base model's spatial understanding while maintaining its other capabilities

## Performance 

### Key Improvements

- VISOR benchmark: +249.6% relative improvement
- T2I-CompBench Spatial: +337.5% relative improvement
- GenEval Position: +1250.0% relative improvement
- Maintains or improves base model's image fidelity (lower FID and CMMD scores than base model)

## Using the Model

See our [GitHub repository][code] to get started.

### Effective Prompting

The model works well with:
- Clear spatial relationship descriptors (left, right, above, below)
- Pairs of distinct objects
- Explicit spatial relationships (e.g., "a photo of A to the right of B")

## Training Details

### Training Data

- Built using the SCOP (Spatial Constraints-Oriented Pairing) data engine
- ~28,000 curated object pairs from COCO
- Enforces criteria for:
  - Visual significance
  - Semantic distinction
  - Spatial clarity
  - Object relationships
  - Visual balance

### Training Process

- Trained for 24,000 steps
- Effective batch size of 4
- Learning rate: 5e-6
- Optimizer: AdamW with β₁=0.9, β₂=0.999
- Weight decay: 1e-2

## Evaluation Results

| Metric | StableDiffusion 1.4 | +CoMPaSS |
|--------|-------------|-----------|
| VISOR uncond (⬆️) | 17.58% | **61.46%** |
| T2I-CompBench Spatial (⬆️) | 0.08 | **0.35** |
| GenEval Position (⬆️) | 0.04 | **0.54** |
| FID (⬇️) | 12.82 | **10.89** |
| CMMD (⬇️) | 0.5548 | **0.3235** |

## Citation

If you use this model in your research, please cite:
```bibtex
@inproceedings{zhang2025compass,
  title={CoMPaSS: Enhancing Spatial Understanding in Text-to-Image Diffusion Models},
  author={Zhang, Gaoyang and Fu, Bingtao and Fan, Qingnan and Zhang, Qi and Liu, Runxing and Gu, Hong and Zhang, Huaqi and Liu, Xinguo},
  booktitle={ICCV},
  year={2025}
}
```

## Contact

For questions about the model, please contact <blurgy@zju.edu.cn>

## Download model

Weights for this model are available in Safetensors format.

[./LICENSE]: <./LICENSE>
[Project page]: <https://compass.blurgy.xyz>
[code]: <https://github.com/blurgyy/CoMPaSS>
[arXiv]: <https://arxiv.org/abs/2412.13195>