HyperFlow: Class-Incremental Semantic Change Detection via Hyperbolic Flow
Authors: Manon Béchaz, Léo Buecher, Emanuele Dalsasso, Ciprian Tomoiaga, Patrick Jayet, Marcin Detyniecki, Laetitia Chapel, Lucas Drumetz, Nicolas Courty, Devis Tuia
Organizations: EPFL, Switzerland · Universit´e de Bretagne Sud, France · IRISA, France · INRIA, France · AXA GDAI Research, Switzerland · Institut Agro Rennes-Angers, France · Lab-STICC, IMT Atlantique, France
Semantic change detection in remote sensing aims to identify and characterize land-cover transitions observed in pairs of satellite or aerial images. Existing methods assume a fixed set of change categories, limiting their applicability in real-world scenarios where changes of interest may vary, for example across geographies or over time. In this work, we tackle the problem of class-incremental semantic change detection, where a model progressively learns to detect new types of semantic change. To do so, we exploit the hierarchical structure of land-cover semantics, where newly introduced categories correspond to fine-grained refinements of existing ones, and represent said semantic categories in hyperbolic space, which is particularly well suited to represent such hierarchical relationships. To capture semantic changes, we introduce HyperFlow, a method relying on Busemann prototypes, modeling change as an anomaly in a learned semantic flow within the hyperbolic embedding space induced by segmentation. HyperFlow couples change detection to class-incremental semantic segmentation, enabling class-incremental semantic change detection. We also present HiChange, a new large-scale, high resolution remote sensing dataset with detailed, fine-grained hierarchical labels. Experimental results demonstrate that HyperFlow achieves both strong change detection and semantic segmentation performance across incremental steps. In particular, we show that progressively refining semantic categories, rather than considering them all directly, yields improved change detection, highlighting the benefits of incremental learning for semantic change detection.
Figures & tables
Figure 1: Overview of CI-SCD via Hyperbolic Flow. (1) Land-cover semantics are naturally organized in hierarchies. (2) During CI-SCD, the model leverages this taxonomy to progressively refine semantic categories from coarse to fine-grained classes. Change maps evolve to account for the progressively refined segmentation. (3) Hyperbolic semantic flow: a vector field transforming the embeddings of the first image into the embeddings of the second image is learned. The hyperbolic distance between the positions of the predicted embeddings and the observed positions indicates change. (4) Prototype-based incremental learning: we refine a coarse class by splitting it into leaves prototype nodes. The vector field in (3) is then retrained to adapt to the new semantic classes.
Figure 2: Qualitative illustration of HyperFlow. (1) SCD at two incremental steps. For each step, we show segmentation results at both timestamps, predicted change map, and a naive baseline obtained by differencing the two segmentation maps (Segm. diff.). (2) Hyperbolic 2- d embeddings of the first and second image, and predicted second embeddings obtained via the learned vector field. Pixels corresponding to true changes (in red) exhibit significant displacement between the timestamps, while unchanged regions show minor displacements. (3) Evolution of the embedding space across incremental steps. We visualize embeddings for a single image before and after the introduction of new classes. More qualitative examples are provided in Supplementary Materials S5 .
HiChange
SECOND
Hi-UCD-mini
Acc.
F1
Pre.
Rec.
Acc.
F1
Pre.
Rec.
Acc.
F1
Pre.
Rec.
CVA ( Malila, 1980 )
74.2
8.0
4.5
39.3
67.6
30.5
25.9
37.1
75.5
14.4
9.3
32.3
SiROC ( Kondmann et al., 2021 )
78.0
9.3
5.3
39.1
69.4
29.6
26.5
33.4
74.4
15.6
9.9
36.9
CDRL ( Noh et al., 2022 )
73.0
11.6
6.4
61.9
62.3
33.2
25.2
48.8
81.6
17.9
12.5
31.3
X-Net ( Luppino et al., 2021 )
69.0
11.5
6.3
70.1
60.8
34.7
25.5
54.2
84.9
20.4
15.4
30.1
ACE-Net ( Luppino et al., 2021 )
69.9
11.0
6.0
65.0
65.7
33.2
26.5
44.4
78.1
20.0
13.1
42.6
Table 1: Comparison of HyperFlow with state-of-the-art unsupervised CD methods across the three datasets, for the task TK00 , i.e . before any incremental step is performed. For HyperFlow, we additionally report CD performance using different embedding sources: HyperFlow uses the in-domain embeddings produced by the segmentation pipeline, while HyperFlow (SAM3, AnySat) use embeddings from the corresponding foundation models. Bold and underline indicate the first and second-best unsupervised scores, respectively. † The HRSCD4 model is supervised.
T40
T71
T132
T163
T204
T225
T326
Segm. [mIoU]
62.0
56.2
33.7
28.4
23.5
20.7
16.1
CD [F1]
29.6
29.6
29.8
29.2
27.7
31.6
26.0
SCD [SCS]
39.0
34.4
22.4
19.9
17.3
17.6
13.8
Table 2: SCD performance of HyperFlow across incremental steps for the HiChange dataset. We report segmentation performance with mIoU, CD with F1, and combined CD and segmentation using the Semantic Change Segmentation (SCS) score from Toker et al. (2022) .
HiChange
SECOND
Hi-UCD
Method
Hyp.
T40
T71
T132
T163
T204
T225
T326
T60
T90
PLOP ( Douillard et al., 2021 )
✗
66.3
11.4
2.0
0.4
0.9
0.8
0.6
53.7
41.8
MiB ( Cermelli et al., 2020 )
✗
66.3
52.9
36.0
27.8
22.2
20.7
16.1
53.7
41.8
MiB+AWT ( Goswami et al., 2023 )
✗
66.3
54.7
37.6
28.2
22.9
20.2
15.9
53.7
41.8
DKD ( Baek et al., 2022 )
✗
64.7
57.4
35.0
27.8
22.1
21.5
17.2
38.8
45.8
TOPICS ( Hindel et al., 2024 )
✓
67.6
54.5
24.6
15.9
8.5
6.2
3.8
46.8
37.5
Table 3: Comparison of Hyperflow with state-of-the-art CISS methods for the three datasets. The mIoU is reported for each method and all incremental steps. Note that since SECOND and Hi-UCD do not enable incremental scenarios, only the base task is considered (Hyp: hyperbolic).
Incr.
Hier.
T40
T71
T132
T163
T204
T225
T326
✓
✓
39.0
34.4
22.4
19.9
17.3
17.6
13.8
✗
✗
39.0
30.9
16.6
13.7
11.7
11.7
9.1
✗
✓
39.0
31.7
17.8
16.2
14.4
14.1
10.0
Table 4: Comparison between (i) class-incremental training with progressive taxonomy refinement, (ii) non-incremental training on the full fine-grained taxonomy, and (iii) non-incremental training with hierarchical prototype initialization for the HiChange dataset. The SCS score is reported for all incremental steps. At each step, the incremental and non incremental models are trained for the same total number of epochs.
Appendix figures & tables21 assets
Supplementary material from the paper’s appendix.
Appendix
Dataset
Image pairs
Image size
Res [m]
Categories
HRSCD Daudt et al. (2019)
291
10000×10000
0.5
5
Hi-UCD-mini Tian et al. (2020)
1293
1024×1024
0.1
9
SECOND Yang et al. (2020)
4662
512×512
not indicated
6
CNAM-CD Zhou et al. (2023)
2503
512×512
0.5
5
DynamicEarthNet Toker et al. (2022)
75 time series of 730 images
1024×1024
3
7
ChangeNet Ji et al. (2024)
31000
∼1900×1200
0.3
6
Appendix
Table 5: Existing SCD datasets, and comparison with the proposed HiChange dataset.
Figure 3: Sampling procedure for constructing the HiChange dataset. (a) Tiles retained after the initial filtering step based on the number of changes observed. (b) Geographical partitioning of French administrative departments used to define train, test and validation splits. (c) Assignment of tiles to the corresponding split according to their geographical location. (d) Final selection of the tiles after the greedy sampling algorithm, ensuring diverse and balanced semantic change types in all splits of the dataset.
Figure 4: Pipeline for the generation of the labels.
Figure 5: Taxonomy of the semantic labels of the HiChange dataset.
Figure 6: Examples from the HiChange dataset, illustrating the refinement of semantic labels. In the initial step, only coarse distinctions between vegetation and constructed areas are considered. At step 1, we have additional information on the type of vegetation and, at step 2, about the type of construction. Note that the change maps adapt to the set of semantic labels considered: in the first example, change happens within the vegetation, and is captured only at steps 1 and 2. Similarly, in the second example, a building is constructed and the change is only captured in the last incremental step.
Figure 7: Distribution of the land-cover categories in the HiChange dataset at different levels of the taxonomy.
Figure 8: Semantic changes in HiChange. Confusion matrices of land-cover transitions for pixels labeled as change at the three hierarchical levels of the taxonomy.
Step
Task
Classes introduced
0
T40
[1,2,3,4]
1
T71
[31,32,33,34]
2
T132
[10,11,12,13,14,15,16]
3
T163
[40,41,42,43]
4
T204
[20,21,22,23,24]
5
T225
[321,322,323]
Appendix
Table 6: Incremental tasks for the HiChange dataset. The table summarizes the classes considered at each incremental step.
Figure 9: Qualitative examples on HiChange of segmentation across incremental steps and associated embedding space visualizations. For each step, we show the predicted segmentation together with the corresponding hyperbolic embeddings, illustrating how the embedding space evolves as the taxonomy is refined and new classes are introduced. The color of each embedding corresponds to the ground-truth semantic label.
Figure 10: Qualitative examples of SCD at task T152 . For each image pair, we show the predicted segmentation at both timesteps, the predicted change map, the embeddings of the two images and the predicted embeddings of the second image. Pixels corresponding to true semantic changes (in red) exhibit larger displacement in the embedding space, that are not captured by the learned flow and therefore not accounted for in the predicted embeddings.
Figure 11: Comparison between incremental and non-incremental scenarios. We visualize the segmentation and change predictions at (a) T132 and (b) T326 , along with corresponding embedding spaces. Incremental refinement preserves the hierarchical organization, enabling better prediction, whereas directly learning fine-grained categories leads to less structured embeddings and lower performances.
Figure 12: Qualitative examples on the SECOND dataset. For each image pair, we show segmentation and change prediction, and the associated embeddings.
Figure 13: Qualitative examples on the Hi-UCD-mini dataset. For each image pair, we show segmentation and change prediction, and the associated embeddings.
Figure 14: Distribution of change scores and triangle threshold for the HiChange dataset and the base task T40 .
Figure 15: Influence of dthresh on CD performance. We present the F1 score of HyperFlow with respect to the threshold used, for HiChange and the base task T40 . Results are compared with the best unsupervised CD baseline, KPCAMNet.
Dataset
Embeddings
Hyperbolic/Euclidean
Threshold
HiChange
Segmentation
Hyp.
4.2
HiChange
Segmentation
Euc.
0.2
HiChange
Segmentation
Euc-Bus.
0.2
HiChange
SAM3
Hyp.
6.3
HiChange
AnySat
Hyp.
3.6
SECOND
Segmentation
Hyp.
7.0
Appendix
Table 7: Thresholds values obtained from the triangle method. Unless stated otherwise, these values are used for all experiments.
Metric
Backbone
T40
T71
T132
T163
T204
T225
T326
Segm. [mIoU]
DeepLabV3
62.0
56.2
33.7
28.4
23.5
20.7
16.1
UNet
50.5
24.3
12.3
9.1
8.0
6.9
4.7
CD [F1]
DeepLabV3
29.6
29.6
29.8
29.2
27.7
31.6
26.0
UNet
14.5
16.3
17.1
17.0
17.2
18.8
17.3
SCD [SCS]
DeepLabV3
39.0
34.4
22.4
19.9
17.3
17.6
13.8
UNet
31.1
16.7
10.3
8.8
8.6
8.5
7.1
Appendix
Table 8: Influence of the backbone architecture on the performances of HyperFlow.
Segmentation prototypes
Trained
Acc.
F1
Pre.
Rec.
✓
✗
92.9
29.2
20.2
52.6
✗
✗
93.1
26.9
19.1
45.5
✗
✓
93.9
27.9
20.8
42.5
Appendix
Table 9: Comparison of segmentation-based and independent prototypes for CD on the HiChange dataset, showing the impact of prototype initialization on HyperFlow performance.
Table 10: Comparison of hyperbolic and Euclidean formulations for HyperFlow on the HiChange dataset and the base task T40 , using (a) triangle thresholds and (b) optimal thresholds.
Dimension
Segm. [mIoU]
CD [F1]
SCD [SCS]
2
62.0
29.6
39.0
16
60.8
21.7
37.1
64
62.2
28.4
39.8
128
58.6
32.6
41.8
Appendix
Table 11: Impact of the embedding dimension d on segmentation, CD, and SCD performance for HiChange and the base task T40 .
Figure 16: Failure cases of HyperFlow. The first two examples illustrate errors caused by inaccuracies in the predicted land-cover, which propagate to the change map. The last two illustrate cases where the ground-truth annotations present errors.
Change detection is a fundamental task in remote sensing, aiming to quantify the impacts of human activities and ecological dynamics on land-cover changes. Existing change detection methods are limited to predefined classes in training datasets, which constrains their scalability in real-world scenarios. In recent years, numerous advanced open-vocabulary semantic segmentation models have emerged for remote sensing imagery. However, there is still a lack of an effective framework for directly applying these models to open-vocabulary change detection (OVCD), a novel task that integrates vision and language to detect changes across arbitrary categories. To address these challenges, we first construct a category-agnostic change detection dataset, termed CA-CDD. Further, we design a category-agnostic change head to detect the transitions of arbitrary categories and index them to specific classes. Based on them, we propose Seg2Change, an adapter designed to adapt open-vocabulary semantic segmentation models to change detection task. Without bells and whistles, this simple yet effective framework achieves state-of-the-art OVCD performance (+9.52 IoU on WHU-CD and +5.50 mIoU on SECOND). Our code is released at https://github.com/yogurts-sy/Seg2Change.
Semantic change detection (SCD) extends the binary change detection task to provide not only the change locations but also the detailed "from-to" categories in multi-temporal remote sensing data. Such detailed semantic insights into changes offer considerable advantages for a wide array of applications. However, since SCD involves the simultaneous optimization of multiple tasks, the model is prone to negative transfer due to task-specific learning difficulties and conflicting gradient flows. To address this issue, we propose Graph-Aggregated Prototype Learning for Remote Sensing Semantic Change Detection(GAPL-SCD). In this framework, a multi-task joint optimization method is designed to optimize the primary task of semantic segmentation and change detection, along with the auxiliary task of graph aggregation prototype learning. Adaptive weight allocation and gradient rotation methods are used to alleviate the conflict between training tasks and improve multi-task learning capabilities. Specifically, the graph aggregation prototype learning module constructs an interaction graph using high-level features. Prototypes serve as class proxies, enabling category-level domain alignment across time points and reducing interference from irrelevant changes. Additionally, the proposed self-query multi-level feature interaction and bi-temporal feature fusion modules further enhance multi-scale feature representation, improving performance in complex scenes. Experimental results on the SECOND and Landsat-SCD datasets demonstrate that our method achieves state-of-the-art performance, with significant improvements in accuracy and robustness for SCD task.
Zhengyi Xu, Haoran Wu, Wen Jiang +1
The authors are with the School of Electronics and Information, Northwestern Polytechnical University, Xi’an, 710129, China.
Remote sensing change detection (RSCD) aims to localise changes between two images of the same geographic region. In practice, change masks often follow region-level annotation conventions rather than purely local appearance differences, making them context-dependent and occasionally ambiguous. Most state-of-the-art methods utilise per-pixel discriminative classification, which produces a single prediction per input and fails to explicitly model the changed region as a coherent whole. A natural alternative is generative formulation, which can model a distribution of plausible masks, enabling sampling to capture ambiguity and encourage global consistency. However, existing generative RSCD approaches typically lag behind strong discriminative baselines due to the high computational cost of pixel-space generation and the complexity of their conditioning mechanisms. To address the limitations of prior discriminative and generative methods, we propose ChangeFlow, a generative framework that reformulates change detection as the synthesis of a change mask in latent space via rectified flow. ChangeFlow is guided by a structured yet lightweight conditioning signal, and its stochastic design naturally supports sampling-based prediction ensembling. Namely, aggregating multiple predicted change masks improves robustness, while sample agreement provides a practical confidence estimation that highlights ambiguous regions. Across four benchmarks, ChangeFlow achieves an average F1 of 80.4%, improving by 1.3 points on average over the previous best method, while maintaining inference speed comparable to recent strong baselines. Project page: https://blaz-r.github.io/changeflow_cd
Blaž Rolih, Matic Fučka, Filip Wolf +1
University of Ljubljana, Faculty of Computer and Information Science, Slovenia