Author name cluster

Andreas Dengel

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

7 papers

1 author row

TMLR Journal 2026 Journal Article

PRISM: Diversifying Dataset Distillation by Decoupling Architectural Priors

Brian Bernhard Moser
Shalini Sarode
Federico Raue
Stanislav Frolov
Krzysztof Adamkiewicz
Arundhati Shanbhag
Joachim Folz
Tobias Christian Nauen

Dataset distillation (DD) promises compact yet faithful synthetic data, but existing approaches often inherit the inductive bias of a single teacher model. As dataset size increases, this bias drives generation toward overly smooth, homogeneous samples, reducing intra-class diversity and limiting generalization. We present PRISM (PRIors from diverse Source Models), a framework that disentangles architectural priors during synthesis. PRISM decouples the logit-matching and regularization objectives, supervising them with different teacher architectures: a primary model for logits and a stochastic subset for batch-normalization (BN) alignment. On ImageNet-1K, PRISM consistently and reproducibly outperforms single-teacher methods (e.g., SRe2L) and recent multi-teacher variants (e.g., G-VBSM) at low- and mid-IPC regimes. The generated data also show significantly richer intra-class diversity, as reflected by a notable drop in cosine similarity between features. We further analyze teacher selection strategies (pre- vs. intra-distillation) and introduce a scalable cross-class batch formation scheme for fast parallel synthesis. Code: https://github.com/Brian-Moser/prism

PDF Details

TMLR Journal 2026 Journal Article

Unifying VXAI: A Systematic Review and Framework for the Evaluation of Explainable AI

David Dembinsky
Adriano Lucieri
Stanislav Frolov
Hiba Najjar
Ko Watanabe
Andreas Dengel

Modern AI systems frequently rely on opaque black-box models, most notably Deep Neural Networks, whose performance stems from complex architectures with millions of learned parameters. While powerful, their complexity poses a major challenge to trustworthiness, particularly due to a lack of transparency. Explainable AI (XAI) addresses this issue by providing human-understandable explanations of model behavior. However, to ensure their usefulness and trustworthiness, such explanations must be rigorously evaluated. Despite the growing number of XAI methods, the field lacks standardized evaluation protocols and consensus on appropriate metrics. To address this gap, we conduct a systematic literature review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and introduce a unified framework for the eValuation of XAI (VXAI). We identify 362 relevant publications and aggregate their contributions into 41 functionally similar metric groups. In addition, we propose a three-dimensional categorization scheme spanning explanation type, evaluation contextuality, and explanation quality desiderata. Our framework provides the most comprehensive and structured overview of VXAI to date. It supports systematic metric selection, promotes comparability across methods, and offers a flexible foundation for future extensions.

PDF Details

NeurIPS Conference 2025 Conference Paper

Unlocking Dataset Distillation with Diffusion Models

Brian Moser
Federico Raue
Sebastian Palacio
Stanislav Frolov
Andreas Dengel

Dataset distillation seeks to condense datasets into smaller but highly representative synthetic samples. While diffusion models now lead all generative benchmarks, current distillation methods avoid them and rely instead on GANs or autoencoders, or, at best, sampling from a fixed diffusion prior. This trend arises because naive backpropagation through the long denoising chain leads to vanishing gradients, which prevents effective synthetic sample optimization. To address this limitation, we introduce Latent Dataset Distillation with Diffusion Models (LD3M), the first method to learn gradient-based distilled latents and class embeddings end-to-end through a pre-trained latent diffusion model. A linearly decaying skip connection, injected from the initial noisy state into every reverse step, preserves the gradient signal across dozens of timesteps without requiring diffusion weight fine-tuning. Across multiple ImageNet subsets at $128\times128$ and $256\times256$, LD3M improves downstream accuracy by up to 4. 8 percentage points (1 IPC) and 4. 2 points (10 IPC) over the prior state-of-the-art. The code for LD3M is provided at https: //github. com/Brian-Moser/prune_and_distill.

PDF Details

AAAI Conference 2022 Conference Paper

Search and Learn: Improving Semantic Coverage for Data-to-Text Generation

Shailza Jolly
Zi Xuan Zhang
Andreas Dengel
Lili Mou

Data-to-text generation systems aim to generate text descriptions based on input data (often represented in the tabular form). A typical system uses huge training samples for learning the correspondence between tables and texts. However, large training sets are expensive to obtain, limiting the applicability of these approaches in real-world scenarios. In this work, we focus on few-shot data-to-text generation. We observe that, while fine-tuned pretrained language models may generate plausible sentences, they suffer from the low semantic coverage problem in the few-shot setting. In other words, important input slots tend to be missing in the generated text. To this end, we propose a search-and-learning approach that leverages pretrained language models but inserts the missing slots to improve the semantic coverage. We further finetune our system based on the search results to smooth out the search noise, yielding better-quality text and improving inference efficiency to a large extent. Experiments show that our model achieves high performance on E2E and WikiBio datasets. Especially, we cover 98. 35% of input slots on E2E, largely alleviating the low coverage problem.

PDF Details

JAIR Journal 2018 Journal Article

Symbol Grounding Association in Multimodal Sequences with Missing Elements

Federico Raue
Andreas Dengel
Thomas M. Breuel
Marcus Liwicki

In this paper, we extend a symbolic association framework for being able to handle missing elements in multimodal sequences. The general scope of the work is the symbolic associations of object-word mappings as it happens in language development in infants. In other words, two different representations of the same abstract concepts can associate in both directions. This scenario has been long interested in Artificial Intelligence, Psychology, and Neuroscience. In this work, we extend a recent approach for multimodal sequences (visual and audio) to also cope with missing elements in one or both modalities. Our method uses two parallel Long Short-Term Memories (LSTMs) with a learning rule based on EM-algorithm. It aligns both LSTM outputs via Dynamic Time Warping (DTW). We propose to include an extra step for the combination with the max operation for exploiting the common elements between both sequences. The motivation behind is that the combination acts as a condition selector for choosing the best representation from both LSTMs. We evaluated the proposed extension in the following scenarios: missing elements in one modality (visual or audio) and missing elements in both modalities (visual and sound). The performance of our extension reaches better results than the original model and similar results to individual LSTM trained in each modality.

PDF Details DOI

EAAI Journal 2014 Journal Article

Intuitive justifications of medical semantic search results

Björn Forcher
Thomas Roth-Berghofer
Stefan Agne
Andreas Dengel

To some extent, explanations in computer science are answers to questions. Often an explanatory dialogue is necessary to satisfy needs of software users. In this paper, we introduce the concept of intuitive explanation representing the first explanations in an explanatory dialogue. This kind of explanation does not require a situational context to be established or that there is a user model. Depending on an abstract model of explanation generation we present the generic explanation component Kalliope applying Semantic Technologies to construct intuitive explanations. We illustrate our generation approach by means of the semantic search engine KOIOS++ enabling keyword-based search on medical articles. Since semantic search results are often hard to understand Kalliope was integrated into KOIOS++ in order to justify search results. In this work we describe in detail the construction of intuitive explanations for inexperienced users in the medical domain building on the concepts of Semantic Frequency Classes and Semantic Cooccurrence Classes. Various user experiments illustrate that these concepts enable the explanation component to rate the understandability of labels and of label connections. We show how Kalliope exploits these valuations to construct and select understandable explanations.

Details DOI

IJCAI Conference 1989 Conference Paper

ANASTASIL: A Hybrid Know ledge-Based System for Document Layout Analysis

Andreas Dengel
Gerhard Barth

This paper describes a know ledge-based system for the identification of the different regions of a document image. It uses a hybrid, modular knowledge representation, a so called geometric tree being its essential part. This tree is used to perform a best-first search in combination with a "hypothesize & test"strategy. It produces an internal, editable description of the entire document and its constituents. The system has been implemented for the analysis of single-sided business letters in Common Lisp on a SUN 3/60 Workstation. It is running for a large population of different business letters. The results obtained have been very encouraging and have convincingly confirmed the soundness of the approach.

PDF