MKMarc A. Kastner

Marc A. Kastner

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Marc A. Kastner has a total of 46 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    MUWS 2026: The 5th International Workshop on Multimodal Human Understanding for the Web and Social Media

    by Eric Müller-Budack, Sherzod Hakimov, Matthias Springstein, Jing Yang, Takahiro Komamizu, Marc A. Kastner

  • Open Access

    Fairness Auditing of Limited-Access Models Using Protected Datasets

    by Zhaohui Zhu, Marc A. Kastner, Shin’iChi Satoh

    Fairness auditing of machine learning models is essential for ensuring equitable treatment across demographic groups, while real-world scenarios often involve strict privacy and access constraints. In this work, we address a practical yet underexplored scenario in which an auditor holds a protected private dataset for evaluation, while the uncooperative target model offers only limited access. Conventional evaluation methods typically rely on either revealing the private dataset or requiring extensive cooperation, which is prevented in the scenario. To overcome this challenge, we propose a novel fairness auditing framework based on model behavioral comparison. Our method constructs a pool of reference models with controlled demographic biases trained on the private dataset, and an embedding space representing fairness similarity. The fairness of the target model is estimated based on the proximity to reference models in the embedding space. We further develop an efficient probe sampling strategy that significantly reduces the number of queries to the target model while maintaining estimation accuracy. We demonstrate the effectiveness of our framework on facial age estimation models and gender classification models, evaluating both gender and racial bias. Evaluation on internally trained models and external pre-trained models is carried out using FairFace and UTKFace datasets to simulate the private dataset and probe data. The experimental results show that our framework achieves close estimation with low mean absolute errors to the ground-truth fairness metrics. Compared to baselines based on annotated public datasets or leaking private data, our framework achieves accurate and efficient fairness evaluation while preserving data confidentiality and operating under limited model access. This framework enables practical, privacy-preserving fairness auditing for real-world scenarios involving regulatory and certification contexts where mutual distrust between auditors and service providers creates operational constraints.

  • Open Access

    Small Object Detection for Birds with Swin Transformer

    by Da Huo, Marc A. Kastner, Tingwei Liu, Yasutomo Kawanishi, Takatsugu Hirayama, Takahiro Komamizu, Ichiro Ide

    Object detection is the task of detecting objects in an image. In this task, the detection of small objects is particularly difficult. Other than the small size, it is also accompanied by difficulties due to blur, occlusion, and so on. Current small object detection methods are tailored to small and dense situations, such as pedestrians in a crowd or far objects in remote sensing scenarios. However, when the target object is small and sparse, there is a lack of objects available for training, making it more difficult to learn effective features. In this paper, we propose a specialized method for detecting a specific category of small objects; birds. Particularly, we improve the features learned by the neck; the sub-network between the backbone and the prediction head, to learn more effective features with a hierarchical design. We employ Swin Transformer to upsample the image features. Moreover, we change the shifted window size for adapting to small objects. Experiments show that the proposed Swin Transformer-based neck combined with CenterNet can lead to good performance by changing the window sizes. We further find that smaller window sizes (default 2) benefit mAPs for small object detection.

  • Open Access

    Semantic Alignment on Action for Image Captioning

    by Da Huo, Marc A. Kastner, Takatsugu Hirayama, Takahiro Komamizu, Yasutomo Kawanishi, Ichiro Ide

    Image captioning is a popular task in vision and language processing, which aims to generate textual descriptions for images. Previously, it simply used image and text as input with self-attention to capture global dependencies. Recent research further uses objects detected from the input image, so-called object tags, as anchor points to ease alignment between image and text with the attention mechanism. However, they only consider object information in images, while neglecting the actions and object interactions that also appear in the image, which causes actions not caught properly in image captioning. To tackle this previously underrepresented dimension of the semantic alignment, we take account of actions on the semantic level. Specifically, our work focuses on human actions and interactions, which ensures that more salient parts of the image get captioned. We introduce a new type of tag, called action tag, to anchor the action information. First, we provide a method for obtaining such action tags using an action detection model which predicts actions in the image. Next, we leverage these action tags into the captioning model. Experimental results indicate that the proposed action tags can help learn action semantics and catch the salient actions leading to perceived improvements in common performance. Experimental results on MS-COCO Karpathy test split show that the proposed model achieves good scores in BLEU-4 and CIDEr metrics, using action tags as anchors. Furthermore, the number of action tags (no more than 5) is smaller than that of object tags (commonly more than 20), which means there is a potential to reduce FLOPs by reducing the total sequence length. It indicates the potential for efficient reasoning and may be applied to daily activity scenes in the future.

  • Lip Shape-Aware Word Selection for Lyric Translation

    by Kotaro Ikeda, Chihaya Matsuhira, Hirotaka Kato, Marc A. Kastner, Takatsugu Hirayama, Takahiro Komamizu, Ichiro Ide

  • Open Access

    MUWS 2025: The 4th International Workshop on Multimodal Human Understanding for the Web and Social Media

    by Sherzod Hakimov, David Semedo, Eric Müller-Budack, Marc A. Kastner, Takahiro Komamizu

  • Open Access

    IntentVC 2025: The ACM Multimedia Grand Challenge on Intention-Oriented Controllable Video Captioning

    by Takahiro Komamizu, Marc A. Kastner, Yasutomo Kawanishi, Trung Thanh Nguyen, Junan Chen


  • Open Access

    Analyzing the Visual Variety of Adjectives based on Clustering of Visual Features

    by Yui Tanaka, Marc A. Kastner, Yasutomo Kawanishi, Takahiro Komamizu, Ichiro Ide

  • Open Access

    ICDAR 25: Intelligent Cross-Data Analysis and Retrieval

    by Takahiro Komamizu, Marc A. Kastner, Minh-Son Dao, Michael Alexander Riegler, Duc-Tien Dang-Nguyen, Son Tran

  • Open Access

    Computer-Assisted Noise Pareidolia Tests Through Patient Emulation

    by Zhaohui Zhu, Marc A. Kastner, Kazuki Nakajima, Shota Furukawa, Momoko Kitazawa, Taishiro Kishimoto, Shin’iChi Satoh

    Pareidolia is a psychiatric phenomenon similar to visual illusions or hallucinations, serving as a distinguishing feature of psychiatric disorders. The so-called noise pareidolia test is a typical diagnosis method for pareidolia. In the test, ambiguous noise-like images are shown to patients, to see if patients misunderstand them as faces. However, the existing testing method is very time-consuming for both patients and experts. Furthermore, it has low accuracy in distinguishing some psychiatric disorders with similar characteristics. This research presents a machine learning-based framework to assist in assessing the pareidolia phenomenon. This is a challenging task because analyzing psychiatric disorders primarily involves observing behavioral changes. This is in contrast to other medical imaging tasks which rely on physically visible changes. Our framework emulates the behavior of patients in noise pareidolia tests. To achieve this, we fine-tune face detection models on patient data to misunderstand ambiguous patterns in the sameway a patient would do. Then, we can compare newpatients to the behavior of existing patients to assess the pareidolia phenomenon, through two classification approaches: a distance function-based approach and a multilayer perceptron-based approach. The proposed framework is evaluated on the data from 48 real patients. The result shows a performance superior to baseline methods in distinguishing pareidolia from other similar visual illusions. Moreover, our framework introduces two sampling methods to reduce the number of images shown to the patient while reaching a comparable high accuracy, improving the efficiency and making the framework feasible for a clinical test.