Diffusion-based world models have demonstrated strong capabilities in synthesizing realistic long-horizon trajectories for offline reinforcement learning (RL). However, many existing methods do not directly generate actions alongside states and rewards, limiting their compatibility with standard value-based offline RL algorithms that rely on one-step temporal difference (TD) learning. While prior work has explored joint modeling of states, rewards, and actions to address this issue, such formulations often lead to increased training complexity and reduced performance in practice. We propose DAWM, a diffusion-based world model that generates future state-reward trajectories conditioned on the current state, action, and return-to-go, paired with an inverse dynamics model (IDM) for efficient action inference. This modular design produces complete synthetic transitions suitable for one-step TD-based offline RL, enabling effective and computationally efficient training. Empirically, we show that conservative offline RL algorithms such as TD3BC and IQL benefit significantly from training on these augmented trajectories, consistently outperforming prior diffusion-based baselines across multiple tasks in the D4RL benchmark.

Segmenting objects in an environment is a crucial task for autonomous driving and robotics, as it enables a better understanding of the surroundings of each agent. Although camera sensors provide rich visual details, they are vulnerable to adverse weather conditions. In contrast, radar sensors remain robust under such conditions, but often produce sparse and noisy data. Therefore, a promising approach is to fuse information from both sensors. In this work, we propose a novel framework to enhance camera-only baselines by integrating a diffusion model into a camera-radar fusion architecture. We leverage radar point features to create pseudo-masks using the Segment-Anything model, treating the projected radar points as point prompts. Additionally, we propose a noise reduction unit to denoise these pseudo-masks, which are further used to generate inpainted images that complete the missing information in the original images. Our method improves the camera-only segmentation baseline by 2.63% in mIoU and enhances our camera-radar fusion architecture by 1.48% in mIoU on the Waterscenes dataset. This demonstrates the effectiveness of our approach for semantic segmentation using camera-radar fusion under adverse weather conditions.

Semantic segmentation represents a fundamental task in computer vision with various application areas such as autonomous driving, medical imaging, or remote sensing. For evaluating and comparing semantic segmentation models, the mean intersection over union (mIoU) is currently the gold standard. However, while mIoU serves as a valuable benchmark, it does not offer insights into the types of errors incurred by a model. Moreover, different types of errors may have different impacts on downstream applications. To address this issue, we propose an intuitive method for the systematic categorization of errors, thereby enabling a fine-grained analysis of semantic segmentation models. Since we assign each erroneous pixel to precisely one error type, our method seamlessly extends the popular IoU-based evaluation by shedding more light on the false positive and false negative predictions. Our approach is model- and dataset-agnostic, as it does not rely on additional information besides the predicted and ground-truth segmentation masks. In our experiments, we demonstrate that our method accurately assesses model strengths and weaknesses on a quantitative basis, thus reducing the dependence on time-consuming qualitative model inspection. We analyze a variety of state-of-the-art semantic segmentation models, revealing systematic differences across various architectural paradigms. Exploiting the gained insights, we showcase that combining two models with complementary strengths in a straightforward way is sufficient to consistently improve mIoU, even for models setting the current state of the art on ADE20K. We release a toolkit for our evaluation method at https://github.com/mxbh/beyond-iou.

Remote sensing projects typically generate large amounts of imagery that can be used to train powerful deep neural networks. However, the amount of labeled images is often small, as remote sensing applications generally require expert labelers. Thus, semi-supervised learning (SSL), i.e., learning with a small pool of labeled and a larger pool of unlabeled data, is particularly useful in this domain. Current SSL approaches generate pseudo-labels from model predictions for unlabeled samples. As the quality of these pseudo-labels is crucial for performance, utilizing additional information to improve pseudo-label quality yields a promising direction. For remote sensing images, geolocation and recording time are generally available and provide a valuable source of information as semantic concepts, such as land cover, are highly dependent on spatiotemporal context, e.g., due to seasonal effects and vegetation zones. In this paper, we propose to exploit spatiotemporal metainformation in SSL to improve the quality of pseudo-labels and, therefore, the final model performance. We show that directly adding the available metadata to the input of the predictor at test time degenerates the prediction quality for metadata outside the spatiotemporal distribution of the training set. Thus, we propose a teacher-student SSL framework where only the teacher network uses metainformation to improve the quality of pseudo-labels on the training set. Correspondingly, our student network benefits from the improved pseudo-labels but does not receive metadata as input, making it invariant to spatiotemporal shifts at test time. Furthermore, we propose methods for encoding and injecting spatiotemporal information into the model and introduce a novel distillation mechanism to enhance the knowledge transfer between teacher and student. Our framework dubbed Spatiotemporal SSL can be easily combined with several stat...

Video Instance Segmentation (VIS) has seen a growing reliance on query propagation-based methods to model complex and lengthy videos. While these methods dominate the performance, they do not explicitly model discrete events, e.g., occlusion, disappearance, and reappearance. Such events often results in degraded object features over time. We believe learning these events end-to-end with the propagation network would prevent the degradation. To this end, we propose a novel propagation method that models these discrete events with a gating mechanism. First, the gate identifies degraded object features caused by these events. Second, we apply a residual configuration to rectify the feature degradation, alleviating the need for a conventional memory bank. Third, we restrict interaction between relevant and degraded objects with a novel gated self-attention. The gated residual configuration and self-attention forms GRAtt block, which can easily be integrated into the existing propagation frameworks. GRAtt-VIS performs on par with the state-of-the-art methods on YTVIS-19,-21,-22 and challenging OVIS datasets by significantly improving performance over previous methods. The code is available in the supplementary.

Finding meaningful groups, i.e., clusters, in high-dimensional data such as images or texts without labeled data at hand is an important challenge in data mining. In recent years, deep clustering methods have achieved remarkable results in these tasks. However, most of these methods require the user to specify the number of clusters in advance. This is a major limitation since the number of clusters is typically unknown if labeled data is unavailable. Thus, an area of research has emerged that addresses this problem. Most of these approaches estimate the number of clusters separated from the clustering process. This results in a strong dependency of the clustering result on the quality of the initial embedding. Other approaches are tailored to specific clustering processes, making them hard to adapt to other scenarios. In this paper, we propose UNSEEN, a general framework that, starting from a given upper bound, is able to estimate the number of clusters. To the best of our knowledge, it is the first method that can be easily combined with various deep clustering algorithms. We demonstrate the applicability of our approach by combining UNSEEN with the popular deep clustering algorithms DCN, DEC, and DKM and verify its effectiveness through an extensive experimental evaluation on several image and tabular datasets. Moreover, we perform numerous ablations to analyze our approach and show the importance of its components. The code is available at: https://github.com/collinleiber/UNSEEN

The traveling officer problem (TOP) is a challenging stochastic optimization task. In this problem, a parking officer is guided through a city equipped with parking sensors to fine as many parking offenders as possible. A major challenge in TOP is the dynamic nature of parking offenses, which randomly appear and disappear after some time, regardless of whether they have been fined. Thus, solutions need to dynamically adjust to currently fineable parking offenses while also planning ahead to increase the likelihood that the officer arrives during the offense taking place. Though various solutions exist, these methods often struggle to take the implications of actions on the ability to fine future parking violations into account. This paper proposes SATOP, a novel spatial-aware deep reinforcement learning approach for TOP. Our novel state encoder creates a representation of each action, leveraging the spatial relationships between parking spots, the agent, and the action. Furthermore, we propose a novel message-passing module for learning future inter-action correlations in the given environment. Thus, the agent can estimate the potential to fine further parking violations after executing an action. We evaluate our method using an environment based on real-world data from Melbourne. Our results show that SATOP consistently outperforms state-of-the-art TOP agents and is able to fine up to 22% more parking offenses.

Explanations play an essential role in helping users evaluate results from recommender systems. Various natural language generation methods have been proposed to generate explanations for the recommendation. However, they usually suffer from two problems. First, since user-provided review text contains noisy data, the generated explanations may be irrelevant to the recommended items. Second, as lacking some supervision signals, most of the generated sentences are similar, which cannot meet the diversity and personalized needs of users. To tackle these problems, we propose a multimodal contrastive transformer (MMCT) model for an explainable recommendation, which incorporates multimodal information into the learning process, including sentiment features, item features, item images, and refined user reviews. Meanwhile, we propose a dynamic fusion mechanism during the decoding stage, which generates supervision signals to guide the explanation generation. Additionally, we develop a contrastive objective to generate diverse explainable texts. Comprehensive experiments on two real-world datasets show that the proposed model outperforms comparable explainable recommendation baselines in terms of explanation performance and recommendation performance. Efficiency analysis and robustness analysis verify the advantages of the proposed model. While ablation analysis establishes the relative contributions of the respective components and various modalities, the case study shows the working of our model from an intuitive sense.

Allocation tasks represent a class of problems where a limited amount of resources must be allocated to a set of entities at each time step. Prominent examples of this task include portfolio optimization or distributing computational workloads across servers. Allocation tasks are typically bound by linear constraints describing practical requirements that have to be strictly fulfilled at all times. In portfolio optimization, for example, investors may be obligated to allocate less than 30\% of the funds into a certain industrial sector in any investment period. Such constraints restrict the action space of allowed allocations in intricate ways, which makes learning a policy that avoids constraint violations difficult. In this paper, we propose a new method for constrained allocation tasks based on an autoregressive process to sequentially sample allocations for each entity. In addition, we introduce a novel de-biasing mechanism to counter the initial bias caused by sequential sampling. We demonstrate the superior performance of our approach compared to a variety of Constrained Reinforcement Learning (CRL) methods on three distinct constrained allocation tasks: portfolio optimization, computational workload distribution, and a synthetic allocation benchmark. Our code is available at: https://github.com/niklasdbs/paspo

Financial portfolio managers typically face multi-period optimization tasks such as short-selling or investing at least a particular portion of the portfolio in a specific industry sector. A common approach to tackle these problems is to use constrained Markov decision process (CMDP) methods, which may suffer from sample inefficiency, hyperparameter tuning, and lack of guarantees for constraint violations. In this paper, we propose Action Space Decomposition Based Optimization (ADBO) for optimizing a more straightforward surrogate task that allows actions to be mapped back to the original task. We examine our method on two real-world data portfolio construction tasks. The results show that our new approach consistently outperforms state-of-the-art benchmark approaches for general CMDPs.

Recent trends in Video Instance Segmentation (VIS) have seen a growing reliance on online methods to model complex and lengthy video sequences. However, the degradation of representation and noise accumulation of the online methods, especially during occlusion and abrupt changes, pose substantial challenges. Transformer-based query propagation provides promising directions at the cost of quadratic memory attention. However, they are susceptible to the degradation of instance features due to the above-mentioned challenges and suffer from cascading effects. The detection and rectification of such errors remain largely underexplored. To this end, we introduce \textbf{GRAtt-VIS}, \textbf{G}ated \textbf{R}esidual \textbf{Att}ention for \textbf{V}ideo \textbf{I}nstance \textbf{S}egmentation. Firstly, we leverage a Gumbel-Softmax-based gate to detect possible errors in the current frame. Next, based on the gate activation, we rectify degraded features from its past representation. Such a residual configuration alleviates the need for dedicated memory and provides a continuous stream of relevant instance features. Secondly, we propose a novel inter-instance interaction using gate activation as a mask for self-attention. This masking strategy dynamically restricts the unrepresentative instance queries in the self-attention and preserves vital information for long-term tracking. We refer to this novel combination of Gated Residual Connection and Masked Self-Attention as \textbf{GRAtt} block, which can easily be integrated into the existing propagation-based framework. Further, GRAtt blocks significantly reduce the attention overhead and simplify dynamic temporal modeling. GRAtt-VIS achieves state-of-the-art performance on YouTube-VIS and the highly challenging OVIS dataset, significantly improving over previous methods. Code is available at \url{https://github.com/Tanveer81/GRAttVIS}.

The task of portfolio management is the selection of portfolio allocations for every single time step during an investment period while adjusting the risk-return profile of the portfolio to the investor’s individual level of risk preference. In practice, it can be hard for an investor to quantify his individual risk preference. As an alternative, approximating the risk-return Pareto front allows for the comparison of different optimized portfolio allocations and hence for the selection of the most suitable risk level. Furthermore, an approximation of the Pareto front allows the analysis of the overall risk sensitivity of various investment policies. In this paper, we propose a deep reinforcement learning (RL) based approach, in which a single meta agent generates optimized portfolio allocation policies for any level of risk preference in a given interval. Our method is more efficient than previous approaches, as it only requires training of a single agent for the full approximate risk-return Pareto front. Additionally, it is more stable in training and only requires per time step market risk estimations independent of the policy. Such risk control per time step is a common regulatory requirement for e.g., insurance companies. We benchmark our meta agent against other state-of-the-art risk-aware RL methods using a realistic environment based on real-world Nasdaq-100 data. Our evaluation shows that the proposed meta agent outperforms various benchmark approaches by generating strategies with better risk-return profiles.

A common problem in Graph Neural Networks (GNNs) is known as over-smoothing. By increasing the number of iterations within the message-passing of GNNs, the nodes' representations of the input graph align and become indiscernible. The latest models employing attention mechanisms with Graph Transformer Layers (GTLs) are still restricted to the layer-wise computational workflow of a GNN that are not beyond preventing such effects. In our work, we relax the GNN architecture by means of implementing a routing heuristic. Specifically, the nodes' representations are routed to dedicated experts. Each expert calculates the representations according to their respective GNN workflow. The definitions of distinguishable GNNs result from k-localized views starting from the central node. We call this procedure Graph Sh ell Attention (SEA), where experts process different subgraphs in a transformer-motivated fashion. Intuitively, by increasing the number of experts, the models gain in expressiveness such that a node's representation is solely based on nodes that are located within the receptive field of an expert. We evaluate our architecture on various benchmark datasets showing competitive results while drastically reducing the number of parameters compared to state-of-the-art models.

Portfolio optimization tasks describe sequential decision problems in which the investor's wealth is distributed across a set of assets. Allocation constraints are used to enforce minimal or maximal investments into particular subsets of assets to control for objectives such as limiting the portfolio's exposure to a certain sector due to environmental concerns. Although methods for constrained Reinforcement Learning (CRL) can optimize policies while considering allocation constraints, it can be observed that these general methods yield suboptimal results. In this paper, we propose a novel approach to handle allocation constraints based on a decomposition of the constraint action space into a set of unconstrained allocation problems. In particular, we examine this approach for the case of two constraints. For example, an investor may wish to invest at least a certain percentage of the portfolio into green technologies while limiting the investment in the fossil energy sector. We show that the action space of the task is equivalent to the decomposed action space, and introduce a new reinforcement learning (RL) approach CAOSD, which is built on top of the decomposition. The experimental evaluation on real-world Nasdaq-100 data demonstrates that our approach consistently outperforms state-of-the-art CRL benchmarks for portfolio optimization.

Quantum Machine Learning (QML) is a recent and rapidly evolving field where the theoretical framework and logic of quantum mechanics are employed to solve machine learning tasks. Various techniques with different levels of quantum-classical hybridization have been proposed. Here we focus on variational quantum circuits (VQC), which emerged as the most promising candidates for the quantum counterpart of neural networks in the noisy intermediate-scale quantum (NISQ) era. Although showing promising results, VQCs can be hard to train because of different issues, e.g., barren plateau, periodicity of the weights, or choice of architecture. This paper focuses on this last problem for finding optimal architectures of variational quantum circuits for various tasks. To address it, we propose a gradient-free algorithm inspired by natural evolution to optimize both the weights and the architecture of the VQC. In particular, we present a version of the well-known neuroevolution of augmenting topologies (NEAT) algorithm and adapt it to the case of variational quantum circuits. We refer to the proposed architecture search algorithm for VQC as QNEAT. We test the algorithm with different benchmark problems of classical fields of machine learning i.e. reinforcement learning and combinatorial optimization.

We evaluate recent developments in deep-learning based Single-Image Super Resolution (SISR) on two multi-spectral datasets in the setting of urban heat analysis. The datasets target a Land Surface Temperature (LST) product and top-of-the-atmosphere (TOA) LWIR radiance, respectively. In doing so, we demonstrate the potential of generative modeling approaches, particularly a Super Resolution Generative Adversarial Network (SRGAN), to increase the spatial resolution of thermal data products. We extend the original SRGAN model with additional bands from the visible optical spectrum to increase the spatial resolution up to four times and estimate the model’s predictive uncertainty. This Multi-Spectral Super Resolution (MSSR) approach yields increases in Peak Signal Noise Ratio (PSNR) of 3dB to 6dB when compared against bicubic upsampling, which is comparable with proceedings in state-of-the-art Multi-Image Super Resolution (MISR) [1]. We further discuss the transferability to other sensors and the limitations of using this approach.

Change detection in remote sensing imagery is essential for a variety of applications such as urban planning, disaster management, and climate research. However, existing methods for identifying semantically changed areas overlook the availability of semantic information in the form of existing maps describing features of the earth's surface. In this paper, we leverage this information for change detection in bi-temporal images. We show that the simple integration of the additional information via concatenation of latent representations suffices to significantly outperform state-of-the-art change detection methods. Motivated by this observation, we propose the new task of *Conditional Change Detection*, where pre-change semantic information is used as input next to bi-temporal images. To fully exploit the extra information, we propose *MapFormer*, a novel architecture based on a multi-modal feature fusion module that allows for feature processing conditioned on the available semantic information. We further employ a supervised, cross-modal contrastive loss to guide the learning of visual representations. Our approach outperforms existing change detection methods by an absolute 11.7% and 18.4% in terms of binary change IoU on DynamicEarthNet and HRSCD, respectively. Furthermore, we demonstrate the robustness of our approach to the quality of the pre-change semantic information and the absence pre-change imagery. The code is available at https://github.com/mxbh/mapformer.

Stochastic Resource Collection (SRC) describes tasks where an agent tries to collect a maximal amount of dynamic resources while navigating through a road network. An instance of SRC is the traveling officer problem (TOP), where a parking officer tries to maximize the number of fined parking violations. In contrast to vehicular routing problems, in SRC tasks, resources might appear and disappear by an unknown stochastic process, and thus, the task is inherently more dynamic. In most applications of SRC, such as TOP, covering realistic scenarios requires more than one agent. However, directly applying multi-agent approaches to SRC yields challenges considering temporal abstractions and inter-agent coordination. In this paper, we propose a novel multi-agent reinforcement learning method for the task of Multi-Agent Stochastic Resource Collection (MASRC). To this end, we formalize MASRC as a Semi-Markov Game which allows the use of temporal abstraction and asynchronous actions by various agents. In addition, we propose a novel architecture trained with independent learning, which integrates the information about collaborating agents and allows us to take advantage of temporal abstractions. Our agents are evaluated on the multiple traveling officer problem, an instance of MASRC where multiple officers try to maximize the number of fined parking violations. Our simulation environment is based on real-world sensor data. Results demonstrate that our proposed agent can beat various state-of-the-art approaches.

Recent transformer-based offline video instance segmentation (VIS) approaches achieve encouraging results and significantly outperform online approaches. However, their reliance on the whole video and the immense computational complexity caused by full Spatio-temporal attention limit them in real-life applications such as processing lengthy videos. In this paper, we propose a single-stage transformer-based efficient online VIS framework named InstanceFormer, which is especially suitable for long and challenging videos. We propose three novel components to model short-term and long-term dependency and temporal coherence. First, we propagate the representation, location, and semantic information of prior instances to model short-term changes. Second, we propose a novel memory cross-attention in the decoder, which allows the network to look into earlier instances within a certain temporal window. Finally, we employ a temporal contrastive loss to impose coherence in the representation of an instance across all frames. Memory attention and temporal coherence are particularly beneficial to long-range dependency modeling, including challenging scenarios like occlusion. The proposed InstanceFormer outperforms previous online benchmark methods by a large margin across multiple datasets. Most importantly, InstanceFormer surpasses offline approaches for challenging and long datasets such as YouTube-VIS-2021 and OVIS. Code is available at https://github.com/rajatkoner08/InstanceFormer.

Dynamic Ambulance Redeployment (DAR) is the task of dynamically assigning ambulances after incidents to base stations to minimize future response times. Though DAR has attracted considerable attention from the research community, existing solutions do not consider using electric ambulances despite the global shift towards electric mobility. In this paper, we are the first to examine the impact of electric ambulances and their required downtime for recharging to DAR and demonstrate that using policies for conventional vehicles can lead to a significant increase in either the number of required ambulances or in the response time to emergencies. Therefore, we propose a new redeployment policy that considers the remaining energy levels, the recharging stations’ locations, and the required recharging time. Our new method is based on minimizing energy deficits (MED) and can provide well-performing redeployment decisions in the novel Dynamic Electric Ambulance Redeployment problem (DEAR). We evaluate MED on a simulation using real-world emergency data from the city of San Francisco and show that MED can provide the required service level without additional ambulances in most cases. For DEAR, MED outperforms various established state-of-the-art solutions for conventional DAR and straightforward solutions to this setting.