🌍 HSACNet: Hierarchical Scale-Aware Consistency Regularized Semi-Supervised Change Detection

The code is coming soon.

---

🔍 Abstract

Semi-supervised change detection (SSCD) aims to detect changes between bi-temporal remote sensing images by utilizing limited labeled data and abundant unlabeled data. Existing methods struggle in complex scenarios, exhibiting poor performance when confronted with noisy data. They typically neglect intra-layer multi-scale features while emphasizing inter-layer fusion, harming the integrity of change objects with different scales. In this paper, we propose HSACNet, a Hierarchical Scale-Aware Consistency regularized Network for SSCD. Specifically, we integrate Segment Anything Model 2 (SAM2), using its Hiera backbone as the encoder to extract inter-layer multi-scale features and applying adapters for parameter-efficient fine-tuning. Moreover, we design a Scale-Aware Differential Attention Module (SADAM) that can precisely capture intra-layer multi-scale change features and suppress noise. Additionally, a dual-augmentation consistency regularization strategy is adopted to effectively utilize the unlabeled data. Extensive experiments across four CD benchmarks demonstrate that our HSACNet achieves state-of-the-art performance, with reduced parameters and computational cost.

---

📊 Results

1. Quantitative Results on WHU-CD and LEVIR-CD

The two numbers in each cell denote the changed-class IoU and overall accuracy (OA), respectively.
Results are reported under semi-supervised settings with varying labeled ratios: 5%, 10%, 20%, 40%.
Bold values indicate state-of-the-art performance.

🔹 WHU-CD

Method	5%	10%	20%	40%
AdvEnt	57.7 / 97.87	60.5 / 97.79	69.5 / 98.50	76.0 / 98.91
s4GAN	57.3 / 97.94	58.0 / 97.81	67.0 / 98.41	74.3 / 98.85
SemiCDNet	56.2 / 97.78	60.3 / 98.02	69.1 / 98.47	70.5 / 98.59
SemiCD	65.8 / 98.37	68.0 / 98.45	74.6 / 98.83	78.0 / 99.01
RC-CD	58.0 / 98.01	61.7 / 98.00	74.0 / 98.83	73.9 / 98.85
SemiPTCD	74.1 / 98.85	74.2 / 98.86	76.9 / 98.95	80.8 / 99.17
UniMatch	78.7 / 99.11	79.6 / 99.11	81.2 / 99.18	83.7 / 99.29
CBFF	79.0 / 99.11	80.5 / 99.15	82.0 / 99.23	82.5 / 99.26
Sup-only	56.4 / 97.90	66.1 / 98.38	74.4 / 98.82	84.3 / 99.34
Ours	81.7 / 99.23	81.3 / 99.20	84.9 / 99.35	86.3 / 99.42

Sup-fully (100% labels): WHU-CD — IoU = 89.5, OA = 99.27

🔹 LEVIR-CD

Method	5%	10%	20%	40%
AdvEnt	67.1 / 98.15	70.8 / 98.38	74.3 / 98.59	75.9 / 98.67
s4GAN	66.6 / 98.16	72.2 / 98.48	75.1 / 98.63	76.2 / 98.68
SemiCDNet	67.4 / 98.11	71.5 / 98.42	74.9 / 98.58	75.5 / 98.63
SemiCD	74.2 / 98.59	77.1 / 98.74	77.9 / 98.79	79.0 / 98.84
RC-CD	74.0 / 98.52	76.1 / 98.65	77.1 / 98.70	77.6 / 98.72
SemiPTCD	71.2 / 98.39	75.9 / 98.65	76.6 / 98.65	77.2 / 98.74
UniMatch	82.1 / 99.03	82.8 / 99.07	82.9 / 99.07	83.0 / 99.08
CBFF	82.1 / 99.03	82.8 / 99.06	83.2 / 99.09	83.3 / 99.08
Sup-only	72.0 / 98.38	77.1 / 98.72	81.1 / 98.96	82.2 / 99.03
Ours	82.2 / 99.03	83.1 / 99.07	83.2 / 99.08	83.5 / 99.10

Sup-fully (100% labels): LEVIR-CD — IoU = 83.8, OA = 99.11

---

2. Quantitative Results on GZ-CD and EGY-BCD

The two numbers in each cell denote the changed-class IoU and overall accuracy (OA), respectively.
Results are reported under semi-supervised settings with varying labeled ratios: 5%, 10%, 20%, 40%.
Bold values indicate state-of-the-art performance.

🔹 GZ-CD

Method	5%	10%	20%	40%
AdvEnt	56.7 / 95.52	57.5 / 95.99	70.3 / 97.28	70.8 / 97.29
s4GAN	59.4 / 96.13	61.6 / 96.23	68.5 / 97.10	69.4 / 97.08
SemiCDNet	57.9 / 95.38	54.9 / 95.52	68.9 / 97.16	69.7 / 97.20
SemiCD	59.5 / 96.27	58.6 / 96.03	67.0 / 97.03	71.5 / 97.36
RC-CD	62.2 / 96.26	63.9 / 96.55	74.1 / 97.69	74.2 / 97.57
UniMatch	68.7 / 97.06	69.5 / 97.41	72.8 / 97.71	71.1 / 97.48
CBFF	67.4 / 97.05	69.3 / 97.24	72.0 / 97.62	76.4 / 97.99
Sup-only	64.8 / 96.51	64.2 / 96.75	69.8 / 97.23	78.2 / 98.08
Ours	72.5 / 97.44	73.1 / 97.64	80.1 / 98.27	81.4 / 98.40

Sup-fully (100% labels): GZ-CD — IoU = 83.3, OA = 98.54

🔹 EGY-BCD

Method	5%	10%	20%	40%
AdvEnt	52.0 / 95.30	58.1 / 96.26	59.8 / 96.46	63.8 / 96.94
s4GAN	53.2 / 95.62	56.5 / 96.26	59.4 / 96.62	64.1 / 96.83
SemiCDNet	52.7 / 95.36	56.9 / 96.02	59.8 / 96.53	63.6 / 96.96
SemiCD	54.3 / 95.79	59.2 / 96.29	61.8 / 96.61	65.4 / 96.96
RC-CD	59.0 / 96.17	61.6 / 96.51	64.6 / 96.79	67.7 / 97.09
UniMatch	62.8 / 96.74	65.5 / 97.10	63.6 / 96.91	67.3 / 97.26
CBFF	63.7 / 96.64	64.3 / 96.95	63.8 / 96.95	67.7 / 97.21
Sup-only	58.2 / 96.05	59.6 / 96.26	65.0 / 96.95	69.1 / 97.35
Ours	68.5 / 97.25	68.1 / 97.28	69.6 / 97.43	70.6 / 97.50

Sup-fully (100% labels): EGY-BCD — IoU = 71.7, OA = 97.62

---

🚀 Getting Started

Pretrained Backbone

Download the pre-trained checkpoint: SAM2.

├── ./pretrained
    ├── sam2_hiera_tiny.pt
    ├── sam2_hiera_small.pt
    ├── sam2_hiera_base_plus.pt
    ├── sam2_hiera_large.pt

Dataset

• WHU-CD: imageA, imageB, and label
• LEVIR-CD: imageA, imageB, and label
• GZ-CD: aistudio
• EGY-BCD: baidu drive, passward:EGYD. google drive

The data directory should follow the structure below:

├── [Your WHU-CD/LEVIR-CD/GZ-CD/EGY-BCD Path]
    ├── A
    ├── B
    └── label

Train/validation/test splits and semi-supervised partitions (e.g., 5%, 10%, 20%, 40% labeled) are provided in the splits/ folder. For example, the structure for splits/whu is:

whu/
├── train.txt
├── val.txt
├── test.txt
├── 5%/
│   ├── labeled.txt
│   └── unlabeled.txt
├── 10%/
│   ├── labeled.txt
│   └── unlabeled.txt
├── 20%/
│   ├── labeled.txt
│   └── unlabeled.txt
└── 40%/
    ├── labeled.txt
    └── unlabeled.txt

---

❤️ Acknowledgements

Thanks to SAM2-UNet and UniMatch for their great work.