CD-FKD: Cross-Domain Feature Knowledge Distillation for Robust Single-Domain Generalization in Object Detection

Junseok Lee; Sungho Shin; Seongju Lee; Kyoobin Lee

cs.CVMar 17, 2026

CD-FKD: Cross-Domain Feature Knowledge Distillation for Robust Single-Domain Generalization in Object Detection

Authors: Junseok Lee, Sungho Shin, Seongju Lee, Kyoobin Lee

Organizations: Junseok Lee is with the Advanced Robotics Lab, LG Electronics, Seoul, Republic of Korea. · Sungho Shin is with the Place AI Flatform, Naver, Gyeonggi-do, Republic of Korea · Seongju Lee and Kyoobin Lee are with the Department of AI Convergence, Gwangju Institute of Science and Technology (GIST), Gwangju, Republic of Korea.

Abstract

Single-domain generalization is essential for object detection, particularly when training models on a single source domain and evaluating them on unseen target domains. Domain shifts, such as changes in weather, lighting, or scene conditions, pose significant challenges to the generalization ability of existing models. To address this, we propose Cross-Domain Feature Knowledge Distillation (CD-FKD), which enhances the generalization capability of the student network by leveraging both global and instance-wise feature distillation. The proposed method uses diversified data through downscaling and corruption to train the student network, whereas the teacher network receives the original source domain data. The student network mimics the features of the teacher through both global and instance-wise distillation, enabling it to extract object-centric features effectively, even for objects that are difficult to detect owing to corruption. Extensive experiments on challenging scenes demonstrate that CD-FKD outperforms state-of-the-art methods in both target domain generalization and source domain performance, validating its effectiveness in improving object detection robustness to domain shifts. This approach is valuable in real-world applications, like autonomous driving and surveillance, where robust object detection in diverse environments is crucial.

Figures & tables

Fig. I: Overview of results of our proposed CD-FKD. The top panel qualitatively compares our method with DivAlign [ 11 ] and Faster R-CNN [ 12 ] , on an Dusk-Rainy. The bottom panel shows a radar chart comparing relative performance.

Fig. II: Illustration of the proposed single-domain generalized object detection using cross-domain FKD. As a KD framework, the frozen teacher network (represented in pink color) receives source domain data, whereas the student network (represented in blue sky color) is provided with downscaled and corrupted source domain data.

Fig. III: Examples of corrupted source domain data

Method	Daytime-Clear	Night-Clear	Dusk-Rainy	Night-Rainy	Daytime-Foggy	Average
Faster R-CNN [ 12 ]	54.9	36.6	27.9	12.1	32.1	27.2
IBN-Net [ 33 ]	49.7	32.1	26.1	14.3	29.6	25.5
SW [ 34 ]	50.6	33.4	26.3	13.7	30.8	26.1
IterNorm [ 35 ]	43.9	29.6	22.8	12.6	28.4	23.4
ISW [ 36 ]	51.3	33.2	25.9	14.1	31.8	26.3
S-DGOD [ 17 ]	56.1	36.6	28.2	16.6	33.5	28.7

TABLE I: Single domain generalization object detection results. Average results are calculated using the results from four target domains to compare the generalization ability. The results of the compared models are obtained from their respective papers. Bold / underline indicates the best/second-best result.

Fig. IV: Qualitative evaluation results of the model’s generalization ability on the Night-Clear, Dusk-Rainy, Night-Rainy, and Daytime-Foggy scenes. The top-row images show the results of Faster R-CNN [ 12 ] . The middle-row images show the results of DivAlign [ 11 ] . The bottom-row images show the results of our method. Red circles and arrows indicate false negatives, while yellow circles and arrows indicate false positives.

Method	bus	bike	car	motor	person	rider	truck	mAP
Faster R-CNN [ 12 ]	37.5	36.6	62.0	14.0	43.6	28.8	41.6	37.7
IBN-Net [ 33 ]	37.8	27.3	49.6	15.1	39.2	27.1	38.9	32.1
SW [ 34 ]	38.7	29.2	49.8	16.6	31.5	28.0	40.2	33.4
IterNorm [ 35 ]	38.5	23.5	38.9	15.8	26.6	25.9	38.1	29.6
ISW [ 36 ]	38.5	28.5	49.6	15.4	31.9	27.5	41.3	33.2
S-DGOD [ 17 ]	40.6	35.1	50.7	19.7	34.7	32.1	43.4	36.6

TABLE II: Quantitative results (%) on the Night-Clear scene.

Method	bus	bike	car	motor	person	rider	truck	mAP
Faster R-CNN [ 12 ]	34.6	22.3	63.6	7.9	24.2	13.2	39.4	29.3
IBN-Net [ 33 ]	37.0	14.8	50.3	11.4	17.9	13.3	38.4	26.1
SW [ 34 ]	35.2	16.7	50.1	10.4	20.1	13.0	38.8	26.3
IterNorm [ 35 ]	32.9	14.1	38.9	11.0	15.5	11.6	35.7	22.8
ISW [ 36 ]	34.7	16.0	50.0	11.1	17.8	12.6	38.8	25.9
S-DGOD [ 17 ]	37.1	19.6	50.9	13.4	19.7	16.3	40.7	28.2

TABLE III: Quantitative results (%) on the Dusk-Rainy scene.

Method	bus	bike	car	motor	person	rider	truck	mAP
Faster R-CNN [ 12 ]	19.8	8.6	30.2	0.3	0.9	4.8	16.6	12.8
IBN-Net [ 33 ]	24.6	10.0	28.4	0.9	8.3	9.8	18.1	14.3
SW [ 34 ]	22.3	7.8	27.6	0.2	10.3	10.0	17.7	13.7
IterNorm [ 35 ]	21.4	6.7	22.0	0.9	9.1	10.6	17.6	12.6
ISW [ 36 ]	22.5	11.4	26.9	0.4	9.9	9.8	17.5	14.1
S-DGOD [ 17 ]	24.4	11.6	29.5	9.8	10.5	11.4	19.2	16.6

TABLE IV: Quantitative results (%) on the Night-Rainy scene.

Method	bus	bike	car	motor	person	rider	truck	mAP
Faster R-CNN [ 12 ]	30.2	26.6	54.9	27.2	35.1	37.6	19.4	33.0
IBN-Net [ 33 ]	29.9	26.1	44.5	24.4	26.2	33.5	22.4	29.6
SW [ 34 ]	30.6	36.2	44.6	25.1	30.7	34.6	23.6	30.8
IterNorm [ 35 ]	29.7	21.8	42.4	24.4	26.0	33.3	21.6	28.4
ISW [ 36 ]	29.5	26.4	49.2	27.9	30.7	34.8	24.0	31.8
S-DGOD [ 17 ]	32.9	28.0	48.8	29.8	32.5	38.2	24.1	33.5

TABLE V: Quantitative results (%) on the Daytime-Foggy scene.

Methods	FKD	DC	NC	DR	NR	DF	Avg.
Faster R-CNN		54.9	36.6	27.9	12.1	32.1	27.2
Ours		58.8	41.1	33.9	14.9	34.0	30.9
Ours	$\mathcal{L}_{\text{glo}}$	62.0	46.3	41.7	21.4	39.5	37.2
Ours	$\mathcal{L}_{\text{ins}}$	62.5	46.7	42.1	21.9	39.7	37.6
Ours	$\mathcal{L}_{\text{glo}}$ + $\mathcal{L}_{\text{ins}}$	62.7	47.3	42.3	23.4	40.2	38.3

TABLE VI: Ablation study results (%) of the proposed CD-FKD. DC, NC, DR, NR, and DF represent Daytime-Clear, Night-Clear, Dusk-Rainy, Night-Rainy, and Daytime-Foggy, respectively.

Fig. V: Heatmap visualization for target domain scenes. The left column displays the original images, the middle column presents the results from Faster R-CNN, and the right column shows the results from our method. The heatmaps highlight the areas where the model focuses, with regions of higher attention marked by a redder hue.

$\alpha$ / $\beta$	DC	NC	DR	NR	DF	Avg.
0.5/1.5	62.5	46.8	42.1	23.1	40.1	38.0
1.0/1.0	62.7	47.3	42.3	23.4	40.2	38.3
1.5/0.5	62.1	46.7	41.8	22.9	40.1	37.9

TABLE VIII: The impact of the hyperparameters

\alpha

and

\beta

\mathcal{L}_{\textit{global}}

and

\mathcal{L}_{\textit{instance}}

in the distillation training phase.

Object Detection

Object detection locates and classifies objects in images or video, balancing localization accuracy, robustness, and inference speed.

See all 624 papers

Generalization Under Scrutiny: Cross-Domain Detection Progresses, Pitfalls, and Persistent Challenges
Apr 9, 2026Saniya M. Deshmukh, Kailash A. Hambarde, Hugo ProençaObject Detection Cross-Domain Generalization
IT: Instituto de Telecomunicações, University of Beira Interior, Covilhã, Portugal.
Prototype-Anchored Generalized Manifold Regression for Unknown-Domain Object Detection
Jul 8, 2026Zihao Zhang, Aming Wu, Yang Li +1Domain Generalization OOD Generalization
College of Intelligence and Computing, Tianjin Key Lab of Machine Learning, Tianjin University, Tianjin 300072, China · School of Computer Science and Information Engineering, Hefei University of Technology, China
VFM $^{4}$ SDG: Unveiling the Power of VFMs for Single-Domain Generalized Object Detection
Apr 23, 2026Yupeng Zhang, Ruize Han, Ningnan Guo +3Detection Transformer Cross-Domain Generalization
College of Intelligence and Computing, Tianjin University, Tianjin 300350, China · Faculty of Computer Science and Artificial Intelligence, Shenzhen University of Advanced Technology, Shenzhen 518107, Guangdong, China

Apr 9, 2026J/K move · Enter open · S save

Generalization Under Scrutiny: Cross-Domain Detection Progresses, Pitfalls, and Persistent Challenges

Saniya M. Deshmukh, Kailash A. Hambarde, Hugo Proença

IT: Instituto de Telecomunicações, University of Beira Interior, Covilhã, Portugal.

Object Detection Cross-Domain Generalization Distribution Shift

Open paper AskPDF arXiv

Object detection models trained on a source domain often exhibit significant performance degradation when deployed in unseen target domains, due to various kinds of variations, such as sensing conditions, environments and data distributions. Hence, regardless the recent breakthrough advances in deep learning-based detection technology, cross-domain object detection (CDOD) remains a critical research area. Moreover, the existing literature remains fragmented, lacking a unified perspective on the structural challenges underlying domain shift and the effectiveness of adaptation strategies. This survey provides a comprehensive and systematic analysis of CDOD. We start upon a problem formulation that highlights the multi-stage nature of object detection under domain shift. Then, we organize the existing methods through a conceptual taxonomy that categorizes approaches based on adaptation paradigms, modeling assumptions, and pipeline components. Furthermore, we analyze how domain shift propagates across detection stages and discuss why adaptation in object detection is inherently more complex than in classification. In addition, we review commonly used datasets, evaluation protocols, and benchmarking practices. Finally, we identify the key challenges and outline promising future research directions. Cohesively, this survey aims to provide a unified framework for understanding CDOD and to guide the development of more robust detection systems.

Abstract

Object Detection

Explore similar work

Generalization Under Scrutiny: Cross-Domain Detection Progresses, Pitfalls, and Persistent Challenges

Prototype-Anchored Generalized Manifold Regression for Unknown-Domain Object Detection

VFM4^{4}4SDG: Unveiling the Power of VFMs for Single-Domain Generalized Object Detection

VFM $^{4}$ SDG: Unveiling the Power of VFMs for Single-Domain Generalized Object Detection