Aerial-Ground Person Re-Identification (AG-ReID) aims to match pedestrian images captured by unmanned aerial vehicles (UAVs) and ground-based surveillance cameras. The task remains highly challenging due to severe viewpoint discrepancies, frequent occlusions, and substantial domain gaps between aerial and ground imagery. Beyond these observable factors, existing methods often overlook two fundamental structural issues: cumulative positional drift during hierarchical feature extraction and semantic inconsistency across cross-view feature distributions. To address these challenges, we propose ProAlign, a Progressive Positional and Prototype-Guided Alignment Network for AG-ReID. ProAlign comprises two key components: a Layer-wise Progressive Positional Embedding (LPPE) module and a View-Aware Prototype Contrastive Learning (PCL) module. LPPE performs adaptive positional calibration throughout transformer layers by employing conditional positional encoding in shallow layers to mitigate local spatial distortions, while introducing learnable static positional embeddings in deeper layers to reinforce global semantic priors. Meanwhile, PCL maintains view-specific aerial and ground prototypes for each identity and enforces cross-view semantic consistency via a dual-prototype contrastive objective. Extensive experiments on CARGO and two real-world AG-ReID benchmarks demonstrate the effectiveness of ProAlign, achieving state-of-the-art overall retrieval performance on CARGO with improvements of 1.60% in Rank-1 accuracy and 5.01% in mAP over the strongest competing methods.
Figures & tables
Figure 1: Illustration of the motivation of ProAlign. (a) Prior methods suffer from cumulative positional drift and cross-view distributional gaps. (b) ProAlign addresses these issues through layer-wise positional calibration (LPPE) and prototype-guided semantic alignment (PCL).
Figure 2: Overview of the ProAlign architecture. Given aerial or ground-view inputs, ProAlign first extracts patch tokens and processes them through transformer layers equipped with LPPE for depth-adaptive spatial correction. In parallel, the PCL module maintains view-specific prototypes to facilitate cross-view identity alignment.
Method
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
SBS He et al. (2023)
50.32
43.09
29.76
72.31
62.99
48.24
67.50
49.73
29.32
31.25
29.00
18.71
PCB Sun et al. (2019)
51.00
44.50
32.20
74.10
67.60
55.10
55.00
44.60
27.00
34.40
30.40
20.10
BoT Luo et al. (2019)
54.81
46.49
32.40
77.68
66.47
51.34
65.00
49.79
29.82
36.25
32.56
21.46
MGN Wang et al. (2018)
54.81
49.08
36.52
83.93
71.05
55.20
65.00
52.96
36.78
31.87
33.47
24.64
VV Kuma et al. (2019) ; Kumar et al. (2020)
45.83
38.84
39.57
72.31
62.99
48.24
67.50
49.73
29.32
31.25
29.00
18.71
Table 1: Performance comparison of the mainstream methods under four settings of the CARGO dataset . “ALL” denotes the overall retrieval performance of each method. “G ↔ G”, “A ↔ A”, and “A ↔ G” represent the performance of each model in several specific retrieval patterns. Rank1, mAP, and mINP are reported (%). The best performance is shown in bold .
Setting
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Baseline
64.10
55.20
41.13
82.14
71.59
58.39
82.50
66.83
50.22
48.12
42.75
29.95
Baseline + LPPE
60.58
54.92
43.13
79.46
74.10
65.12
75.00
63.62
47.28
53.19
52.63
44.73
Baseline + PCL
65.38
58.55
46.49
79.46
73.99
64.52
82.50
67.66
51.20
58.51
57.65
49.68
Baseline + LPPE + PCL
68.27
62.47
51.39
84.82
80.22
72.51
85.00
70.05
54.66
60.64
58.44
49.86
Table 2: Ablation study of the different components in ProAlign on the CARGO dataset. “LPPE” denotes the Layer-wise Positional Embedding module. “PCL” refers to the Prototype Contrastive Learning module. The best performance per metric is shown in bold .
Setting
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Depths of CPE
First 3 layers
68.59
60.97
48.7
85.71
77.74
67.86
80.00
68.32
53.99
60.64
60.31
52.22
First 6 layers
68.27
62.47
51.39
84.82
80.22
72.51
85.00
70.05
54.66
60.64
58.44
49.86
First 9 layers
66.35
60.40
48.43
81.25
76.22
67.17
82.50
68.73
53.71
62.77
61.66
53.87
All layers
67.63
58.87
45.41
83.93
74.92
64.11
80.00
67.58
52.16
61.70
59.60
50.12
Table 3: Performance comparison under different depths of CPE in LPPE. Results are reported under four matching protocols on the CARGO dataset. The best values per column are shown in bold .
Setting
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
γ=0
66.67
57.24
42.85
82.14
73.79
62.44
80.00
64.44
46.03
63.83
57.99
46.82
γ=0.25
67.31
58.54
45.12
82.14
75.25
65.02
85.00
68.98
51.75
62.77
60.54
51.29
γ=0.5
68.27
62.47
51.39
84.82
80.22
72.51
85.00
70.05
54.66
60.64
58.44
49.86
γ=0.75
66.67
58.54
44.67
82.14
73.25
61.65
82.50
68.39
53.43
59.57
58.76
49.85
Table 4: Impact of the cross-view alignment weight γ in PCL. Results are reported under four matching protocols on the CARGO dataset. The best values per column are shown in bold .
Setting
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Different variants of prototype
Inner-Batch
62.82
56.12
43.55
77.68
72.57
62.64
77.50
64.44
48.94
60.64
58.50
49.49
Global Memory
61.86
55.40
42.71
77.68
71.83
61.10
77.50
64.53
48.89
56.38
55.80
47.80
Classifier Weights
62.82
56.30
43.98
81.25
75.06
65.08
75.00
63.90
47.79
53.19
53.52
45.54
View-Aware PCL
68.27
62.47
51.39
84.82
80.22
72.51
85.00
70.05
54.66
60.64
58.44
49.86
Table 5: Comparison between different variants of prototype. Results are reported under four matching protocols on the CARGO dataset. The best values per column are shown in bold .
Setting
Params (M)
GFLOPs
Latency (ms)
Δ Params
Δ Latency
VDT
85.97
60.75
11.19
–
–
GSAlign
111.67
76.92
11.44
+29.9%
+2.23%
ProAlign
86.98
74.69
11.25
+1.17%
+0.54%
Table 6: Computational complexity comparison. Latency is measured as the average forward pass time over 200 runs.
Layer
3
4
5
6
7
8
9
10
11
VDT
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
LPPE-0.25
0.982
0.982
0.977
0.951
0.881
0.851
0.864
0.857
0.986
LPPE-0.5
0.983
0.982
0.960
0.868
0.809
0.839
0.837
0.865
0.931
LPPE-0.75
0.982
0.985
0.958
0.861
0.817
0.842
0.860
0.881
0.941
LPPE-1.0
0.982
0.982
0.973
0.951
0.871
0.850
0.862
0.845
0.974
Table 7: Normalized [CLS]-attention entropy H(l) per layer on CARGO. 1.0 denotes uniform attention over all 128 patches; lower values indicate sharper spatial focus. LPPE- r applies CPE to the first 12r layers (LPPE-0.5 is the default ProAlign setting).
Figure 3: t-SNE visualization of feature distributions. △ and ◯ denote aerial and ground features, respectively. Baseline shows view-dependent clustering with fragmented identity distributions, while ProAlign produces more compact identity-wise clusters and reduces cross-view discrepancy.
Appendix figures & tables4 assets
Supplementary material from the paper’s appendix.
Appendix
Figure 4: Attention map comparison between baseline and ProAlign. In both aerial and ground views, the baseline attention is scattered due to severe viewpoint distortion, while LPPE produces more focused responses on the person.
Setting
Protocol 1: A ↔ G
Protocol 2: G ↔ A
Rank1
mAP
mINP
Rank1
mAP
mINP
SBS [ 38 ]
73.54
59.77
-
73.70
62.27
-
BoT [ 34 ]
70.01
55.47
-
71.20
58.83
-
OSNet [ 41 ]
72.59
58.32
-
74.22
60.99
-
VV [ 39 , 40 ]
77.22
67.23
41.43
79.73
69.83
42.37
Explain [ 4 ]
81.47
72.61
-
82.85
73.39
-
Appendix
Table 8: Performance comparison of the mainstream methods under two settings of the AG-ReID dataset. Results are reported under matching protocols. The best values per column are shown in bold .
Setting
Protocol 1: A ↔ G
Protocol 2: G ↔ A
Protocol 3: A ↔ W
Protocol 4: W ↔ A
Rank1
mAP
Rank1
mAP
Rank1
mAP
Rank1
mAP
BoT [ 34 ]
85.40
77.03
84.65
75.90
89.77
80.48
84.65
76.90
Explain [ 4 ]
87.70
79.00
87.35
78.24
93.67
83.14
87.73
79.08
VDT [ 6 ]
86.46
79.13
86.14
78.12
90.00
82.21
85.26
78.52
AG-ReIDv2 [ 10 ]
88.77
80.72
87.86
78.51
93.62
84.85
88.61
80.11
SeCap [ 30 ]
88.12
80.84
88.24
79.99
91.44
84.01
87.56
80.15
Appendix
Table 9: Performance comparison of the mainstream methods under four settings of the AG-ReID v2 dataset. Results are reported under four matching protocols. The best values per column are shown in bold .
Setting
Protocol 1: ALL
Protocol 2: G ↔ G
Protocol 3: A ↔ A
Protocol 4: A ↔ G
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
Rank1
mAP
mINP
m=0.1
67.63
60.75
47.97
81.25
75.57
66.16
80.00
67.50
51.37
67.02
62.61
52.69
m=0.2
68.27
62.47
51.39
84.82
80.22
72.51
85.00
70.05
54.66
60.64
58.44
49.86
m=0.3
67.31
59.24
45.95
83.93
75.79
65.32
82.50
67.00
49.45
64.89
61.42
51.48
Appendix
Table 10: Impact of the momentum coefficient ( m ) for prototype update in PCL. Results are reported under four matching protocols on the CARGO dataset. The best values per column are shown in bold .
Aerial-Ground Person Re-Identification (AGPReID) remains highly challenging due to drastic viewpoint variations between drones and fixed cameras. Existing methods typically follow a view-invariant paradigm, aligning shared features across views to achieve robustness. However, view-invariant inherently enforces part-level alignment, which ignores view-specific cues and discriminative identity information. To this end, this work proposes ViSA (View-aware Semantic Alignment), a view-aware framework that achieves cross-view semantic consistency containing an Expert-driven Token Generation Module (ETGM) and a Dual-branch Local Fusion Module (DLFM). Technically, the former constructs a set of view-aware experts to generate adaptive semantic queries that perceive viewpoint-specific patterns, while the latter leverages graph reasoning to extract and align local regions responsive to different experts. Extensive experiments on three AGPReID benchmarks including AG-ReID.v2, CARGO and LAGPeR demonstrate that ViSA consistently achieves superior performance, with a notable 10.06% mAP improvement on the challenging CARGO cross-view protocol. The code is available at https://github.com/Cat-Zero/ViSA.
Quan Zhang, Zeqiang Cai, Peiming Zhao +4
Sun Yat-sen University, China · Pazhou Lab (HuangPu), Guangdong, China · Guangdong Province Key Laboratory of Information Security Technology, Guangzhou, China +1
Aerial-Ground Person Re-IDentification (AG-ReID) aims to retrieve the same person across heterogeneous aerial and ground camera platforms. Although great progress has been made, existing methods remain suboptimal due to the direct feature alignment across views, overlooking view-specific cues. To address this issue, we propose a novel Hierarchical Hyperbolic Representation (HiHR) framework for AG-ReID. More specifically, we first extract multi-granularity features based on pre-trained visual-text encoders. Then, we propose a Text-guided Multi-granularity Fusion (TMF) to fuse multi-granularity features and enhance the representation ability of identity features. Furthermore, we introduce the Hierarchical Hyperbolic Learning (HHL) to construct a hierarchical feature structure in a hyperbolic space. This hierarchy includes a coarse level that ensures identity separability and cross-view consistency, and a fine level that preserves view-specific discriminative cues. As a result, our proposed framework can effectively aggregate view-invariant and view-specific discriminative features for AG-ReID. Extensive experiments on four AG-ReID benchmarks demonstrate the effectiveness of our framework. The source code is available at https://github.com/YangQiWei3/HiHR.
Aerial-ground person re-identification is a challenging task due to cross-platform viewpoint variations, which cause severe occlusion and geometric deformation. Existing methods attempt to learn view-invariant representations exclusively within the 2D image space, where drastic viewpoint variations cause the learned features to remain coupled with viewpoint bias. To address this, we propose VR3D, a View-Robust 3D Representation Learning framework that maps images into a unified 3D coordinate space to achieve view-independent feature interaction. Specifically, we introduce View-Robust 3D Representation Interaction, which leverages 3D priors extracted from single 2D observations to lift 2D appearance features into a canonical 3D space. VR3I employs 3D Geometry-Semantic Attention to establish interactions between 2D patches and 3D voxels from corresponding body parts based on their 3D spatial locations, effectively grounding 2D semantics within a 3D framework. In addition, as the reliability of these representations varies across samples due to viewpoint changes and 3D reconstruction errors, we introduce Reliability-Aware Fusion, which estimates sample-specific reliability and adaptively aggregates the multi-source representations. Extensive experiments on three benchmark datasets (CARGO, AG-ReID.v1, and AG-ReID.v2) demonstrate that VR3D outperforms recent methods. For example, it achieves a 5.63% improvement in Rank-1 on CARGO. Our code will be released.
Chao Ji, Shiyu Xuan, Zechao Li
School of Computer Science and Engineering, Nanjing University of Science and Technology