Search

Traditional drug discovery is a long, expensive, and complex process. Advances in Artificial Intelligence (AI) and Machine Learning (ML) are beginning to change this narrative. Here, we provide a comprehensive overview of different AI and ML tools that can be used to streamline and accelerate the drug discovery process. By using data sets to train ML algorithms, it is possible to discover drugs or drug-like compounds relatively quickly, and efficiently. Additionally, we address limitations in AI-based drug discovery and development, including the scarcity of high-quality data to train AI models and ethical considerations. The growing impact of AI on the pharmaceutical industry is also highlighted. Finally, we discuss how AI and ML can expedite the discovery of new antibiotics to combat the problem of worldwide antimicrobial resistance (AMR)., Comment: 65 pages

For efficient and safe autonomous driving, it is essential that autonomous vehicles can predict the motion of other traffic agents. While highly accurate, current motion prediction models often impose significant challenges in terms of training resource requirements and deployment on embedded hardware. We propose a new efficient motion prediction model, which achieves highly competitive benchmark results while training only a few hours on a single GPU. Due to our lightweight architectural choices and the focus on reducing the required training resources, our model can easily be applied to custom datasets. Furthermore, its low inference latency makes it particularly suitable for deployment in autonomous applications with limited computing resources., Comment: Accepted to IROS 2024

Graph Neural Networks (GNNs) excel in handling graph-structured data but often underperform in link prediction tasks compared to classical methods, mainly due to the limitations of the commonly used Message Passing GNNs (MPNNs). Notably, their ability to distinguish non-isomorphic graphs is limited by the 1-dimensional Weisfeiler-Lehman test. Our study presents a novel method to enhance the expressivity of GNNs by embedding induced subgraphs into the graph Laplacian matrix's eigenbasis. We introduce a Learnable Lanczos algorithm with Linear Constraints (LLwLC), proposing two novel subgraph extraction strategies: encoding vertex-deleted subgraphs and applying Neumann eigenvalue constraints. For the former, we conjecture that LLwLC establishes a universal approximator, offering efficient time complexity. The latter focuses on link representations enabling differentiation between $k$-regular graphs and node automorphism, a vital aspect for link prediction tasks. Our approach results in an extremely lightweight architecture, reducing the need for extensive training datasets. Empirically, our method improves performance in challenging link prediction tasks across benchmark datasets, establishing its practical utility and supporting our theoretical findings. Notably, LLwLC achieves 20x and 10x speedup by only requiring 5% and 10% data from the PubMed and OGBL-Vessel datasets while comparing to the state-of-the-art.

Accurate 3D object detection in LiDAR point clouds is crucial for autonomous driving systems. To achieve state-of-the-art performance, the supervised training of detectors requires large amounts of human-annotated data, which is expensive to obtain and restricted to predefined object categories. To mitigate manual labeling efforts, recent unsupervised object detection approaches generate class-agnostic pseudo-labels for moving objects, subsequently serving as supervision signal to bootstrap a detector. Despite promising results, these approaches do not provide class labels or generalize well to static objects. Furthermore, they are mostly restricted to data containing multiple drives from the same scene or images from a precisely calibrated and synchronized camera setup. To overcome these limitations, we propose a vision-language-guided unsupervised 3D detection approach that operates exclusively on LiDAR point clouds. We transfer CLIP knowledge to classify point clusters of static and moving objects, which we discover by exploiting the inherent spatio-temporal information of LiDAR point clouds for clustering, tracking, as well as box and label refinement. Our approach outperforms state-of-the-art unsupervised 3D object detectors on the Waymo Open Dataset ($+23~\text{AP}_{3D}$) and Argoverse 2 ($+7.9~\text{AP}_{3D}$) and provides class labels not solely based on object size assumptions, marking a significant advancement in the field., Comment: Accepted to BMVC 2024

Commutator blueprints can be seen as blueprints for constructing RGD-systems over $\mathbb{F}_2$ with prescribed commutation relations. In this paper we construct several families of Weyl-invariant commutator blueprints, mostly of universal type. Together with the main result of \cite{BiRGD} we obtain new examples of exotic RGD-systems of universal type over $\mathbb{F}_2$., Comment: 29 pages

In this paper we prove that an RGD-system over $\mathbb{F}_2$ with prescribed commutation relations exists if and only if the commutation relations are Weyl-invariant and can be realized in the group $U_+$. This result gives us a machinery to produce new examples of RGD-systems with complicated commutation relations. We also discuss some applications of this result., Comment: 30 pages

We present the Keras domain packages KerasCV and KerasNLP, extensions of the Keras API for Computer Vision and Natural Language Processing workflows, capable of running on either JAX, TensorFlow, or PyTorch. These domain packages are designed to enable fast experimentation, with a focus on ease-of-use and performance. We adopt a modular, layered design: at the library's lowest level of abstraction, we provide building blocks for creating models and data preprocessing pipelines, and at the library's highest level of abstraction, we provide pretrained ``task" models for popular architectures such as Stable Diffusion, YOLOv8, GPT2, BERT, Mistral, CLIP, Gemma, T5, etc. Task models have built-in preprocessing, pretrained weights, and can be fine-tuned on raw inputs. To enable efficient training, we support XLA compilation for all models, and run all preprocessing via a compiled graph of TensorFlow operations using the tf.data API. The libraries are fully open-source (Apache 2.0 license) and available on GitHub., Comment: Submitted to Journal of Machine Learning Open Source Software

Understanding human behavior fundamentally relies on accurate 3D human pose estimation. Graph Convolutional Networks (GCNs) have recently shown promising advancements, delivering state-of-the-art performance with rather lightweight architectures. In the context of graph-structured data, leveraging the eigenvectors of the graph Laplacian matrix for positional encoding is effective. Yet, the approach does not specify how to handle scenarios where edges in the input graph are missing. To this end, we propose a novel positional encoding technique, PerturbPE, that extracts consistent and regular components from the eigenbasis. Our method involves applying multiple perturbations and taking their average to extract the consistent and regular component from the eigenbasis. PerturbPE leverages the Rayleigh-Schrodinger Perturbation Theorem (RSPT) for calculating the perturbed eigenvectors. Employing this labeling technique enhances the robustness and generalizability of the model. Our results support our theoretical findings, e.g. our experimental analysis observed a performance enhancement of up to $12\%$ on the Human3.6M dataset in instances where occlusion resulted in the absence of one edge. Furthermore, our novel approach significantly enhances performance in scenarios where two edges are missing, setting a new benchmark for state-of-the-art.

Let $(W, S)$ be a Coxeter system of rank $n$ and let $p_{(W, S)}(t)$ be its growth function. It is known that $p_{(W, S)}(q^{-1}) < \infty$ holds for all $n \leq q \in \mathbb{N}$. In this paper we will show that this still holds for $q = n-1$, if $(W, S)$ is $2$-spherical. Moreover, we will prove that $p_{(W, S)}(q^{-1}) = \infty$ holds for $q = n-2$, if the Coxeter diagram of $(W, S)$ is the complete graph. These two results provide a complete characterization of the finiteness of the growth function in the case of $2$-spherical Coxeter systems with complete Coxeter diagram., Comment: 11 pages

State-of-the-art Multiple Object Tracking (MOT) approaches have shown remarkable performance when trained and evaluated on current benchmarks. However, these benchmarks primarily consist of clear weather scenarios, overlooking adverse atmospheric conditions such as fog, haze, smoke and dust. As a result, the robustness of trackers against these challenging conditions remains underexplored. To address this gap, we introduce physics-based volumetric fog simulation method for arbitrary MOT datasets, utilizing frame-by-frame monocular depth estimation and a fog formation optical model. We enhance our simulation by rendering both homogeneous and heterogeneous fog and propose to use the dark channel prior method to estimate atmospheric light, showing promising results even in night and indoor scenes. We present the leading benchmark MOTChallenge (third release) augmented with fog (smoke for indoor scenes) of various intensities and conduct a comprehensive evaluation of MOT methods, revealing their limitations under fog and fog-like challenges.

We propose a novel approach to video anomaly detection: we treat feature vectors extracted from videos as realizations of a random variable with a fixed distribution and model this distribution with a neural network. This lets us estimate the likelihood of test videos and detect video anomalies by thresholding the likelihood estimates. We train our video anomaly detector using a modification of denoising score matching, a method that injects training data with noise to facilitate modeling its distribution. To eliminate hyperparameter selection, we model the distribution of noisy video features across a range of noise levels and introduce a regularizer that tends to align the models for different levels of noise. At test time, we combine anomaly indications at multiple noise scales with a Gaussian mixture model. Running our video anomaly detector induces minimal delays as inference requires merely extracting the features and forward-propagating them through a shallow neural network and a Gaussian mixture model. Our experiments on five popular video anomaly detection benchmarks demonstrate state-of-the-art performance, both in the object-centric and in the frame-centric setup.

Alzheimer's disease (AD) is a progressive neurodegenerative disorder leading to cognitive decline. [$^{18}$F]-Fluorodeoxyglucose positron emission tomography ([$^{18}$F]-FDG PET) is used to monitor brain metabolism, aiding in the diagnosis and assessment of AD over time. However, the feasibility of multi-time point [$^{18}$F]-FDG PET scans for diagnosis is limited due to radiation exposure, cost, and patient burden. To address this, we have developed a predictive image-to-image translation (I2I) model to forecast future [$^{18}$F]-FDG PET scans using baseline and year-one data. The proposed model employs a convolutional neural network architecture with long-short term memory and was trained on [$^{18}$F]-FDG PET data from 161 individuals from the Alzheimer's Disease Neuroimaging Initiative. Our I2I network showed high accuracy in predicting year-two [18F]-FDG PET scans, with a mean absolute error of 0.031 and a structural similarity index of 0.961. Furthermore, the model successfully predicted PET scans up to seven years post-baseline. Notably, the predicted [$^{18}$F]-FDG PET signal in an AD-susceptible meta-region was highly accurate for individuals with mild cognitive impairment across years. In contrast, a linear model was sufficient for predicting brain metabolism in cognitively normal and dementia subjects. In conclusion, both the I2I network and the linear model could offer valuable prognostic insights, guiding early intervention strategies to preemptively address anticipated declines in brain metabolism and potentially to monitor treatment effects., Comment: Accepted to SPIE Medical Imaging 2024

A better understanding of interactive pedestrian behavior in critical traffic situations is essential for the development of enhanced pedestrian safety systems. Real-world traffic observations play a decisive role in this, since they represent behavior in an unbiased way. In this work, we present an approach of how a subset of very considerable pedestrian-vehicle interactions can be derived from a camera-based observation system. For this purpose, we have examined road user trajectories automatically for establishing temporal and spatial relationships, using 110h hours of video recordings. In order to identify critical interactions, our approach combines the metric post-encroachment time with a newly introduced motion adaption metric. From more than 11,000 reconstructed pedestrian trajectories, 259 potential scenarios remained, using a post-encroachment time threshold of 2s. However, in 95% of cases, no adaptation of the pedestrian behavior was observed due to avoiding criticality. Applying the proposed motion adaption metric, only 21 critical scenarios remained. Manual investigations revealed that critical pedestrian vehicle interactions were present in 7 of those. They were further analyzed and made publicly available for developing pedestrian behavior models3. The results indicate that critical interactions in which the pedestrian perceives and reacts to the vehicle at a relatively late stage can be extracted using the proposed method., Comment: 7 pages, 8 figures, ITSC 2023 accepted

Widely-used LiDAR-based 3D object detectors often neglect fundamental geometric information readily available from the object proposals in their confidence estimation. This is mostly due to architectural design choices, which were often adopted from the 2D image domain, where geometric context is rarely available. In 3D, however, considering the object properties and its surroundings in a holistic way is important to distinguish between true and false positive detections, e.g. occluded pedestrians in a group. To address this, we present GACE, an intuitive and highly efficient method to improve the confidence estimation of a given black-box 3D object detector. We aggregate geometric cues of detections and their spatial relationships, which enables us to properly assess their plausibility and consequently, improve the confidence estimation. This leads to consistent performance gains over a variety of state-of-the-art detectors. Across all evaluated detectors, GACE proves to be especially beneficial for the vulnerable road user classes, i.e. pedestrians and cyclists., Comment: ICCV 2023, code is available at https://github.com/dschinagl/gace

University research groups in Computational Science and Engineering (CSE) generally lack dedicated funding and personnel for Research Software Engineering (RSE), which, combined with the pressure to maximize the number of scientific publications, shifts the focus away from sustainable research software development and reproducible results. The neglect of RSE in CSE at University research groups negatively impacts the scientific output: research data - including research software - related to a CSE publication cannot be found, reproduced, or re-used, different ideas are not combined easily into new ideas, and published methods must very often be re-implemented to be investigated further. This slows down CSE research significantly, resulting in considerable losses in time and, consequentially, public funding. We propose a RSE workflow for Computational Science and Engineering (CSE) that addresses these challenges, that improves the quality of research output in CSE. Our workflow applies established software engineering practices adapted for CSE: software testing, result visualization, and periodical cross-linking of software with reports/publications and data, timed by milestones in the scientific publication process. The workflow introduces minimal work overhead, crucial for university research groups, and delivers modular and tested software linked to publications whose results can easily be reproduced. We define research software quality from a perspective of a pragmatic researcher: the ability to quickly find the publication, data, and software related to a published research idea, quickly reproduce results, understand or re-use a CSE method, and finally extend the method with new research ideas., Comment: arXiv admin note: text overlap with arXiv:2208.07460

Neural Image Classifiers are effective but inherently hard to interpret and susceptible to adversarial attacks. Solutions to both problems exist, among others, in the form of counterfactual examples generation to enhance explainability or adversarially augment training datasets for improved robustness. However, existing methods exclusively address only one of the issues. We propose a unified framework leveraging image-to-image translation Generative Adversarial Networks (GANs) to produce counterfactual samples that highlight salient regions for interpretability and act as adversarial samples to augment the dataset for more robustness. This is achieved by combining the classifier and discriminator into a single model that attributes real images to their respective classes and flags generated images as "fake". We assess the method's effectiveness by evaluating (i) the produced explainability masks on a semantic segmentation task for concrete cracks and (ii) the model's resilience against the Projected Gradient Descent (PGD) attack on a fruit defects detection problem. Our produced saliency maps are highly descriptive, achieving competitive IoU values compared to classical segmentation models despite being trained exclusively on classification labels. Furthermore, the model exhibits improved robustness to adversarial attacks, and we show how the discriminator's "fakeness" value serves as an uncertainty measure of the predictions.

Vision and Language Models (VLMs), such as CLIP, have enabled visual recognition of a potentially unlimited set of categories described by text prompts. However, for the best visual recognition performance, these models still require tuning to better fit the data distributions of the downstream tasks, in order to overcome the domain shift from the web-based pre-training data. Recently, it has been shown that it is possible to effectively tune VLMs without any paired data, and in particular to effectively improve VLMs visual recognition performance using text-only training data generated by Large Language Models (LLMs). In this paper, we dive deeper into this exciting text-only VLM training approach and explore ways it can be significantly further improved taking the specifics of the downstream task into account when sampling text data from LLMs. In particular, compared to the SOTA text-only VLM training approach, we demonstrate up to 8.4% performance improvement in (cross) domain-specific adaptation, up to 8.7% improvement in fine-grained recognition, and 3.1% overall average improvement in zero-shot classification compared to strong baselines., Comment: Code is available at: https://github.com/jmiemirza/TAP

In the development of locally resonant metamaterials, the physical resonator design is often omitted and replaced by an idealized mass-spring system. This paper presents a novel approach for designing multimodal resonant structures, which give rise to multi-bandgap metamaterials with predefined band gaps. Our method uses a conditional variational autoencoder to identify nontrivial patterns between design variables of complex-shaped resonators and their modal effective parameters. After training, the cost of generating designs satisfying arbitrary criteria - frequency and mass of multiple modes - becomes negligible. An example of a resonator family with six geometric variables and two targeted modes is further elaborated. We find that the autoencoder performs well even when trained with a limited dataset, resulting from a few hundred numerical modal analyses. The method generates several designs that very closely approximate the desired modal characteristics. The accuracy of the best designs, proposed by the auto-encoder, is confirmed in tests of 3D-printed resonator prototypes. Further experiments demonstrate the close agreement between the measured and desired dispersion relation of a sample metamaterial beam.

We compute the disk amplitude of three closed strings in the pure spinor formalism. Among others, this amplitude probes tree-level gravitational interactions in the presence of Dp-branes. After disentangling holomorphic and anti-holomorphic closed string coordinates on the disk by means of introducing monodromy phases we find a compact expression for the disk amplitude of three closed strings in terms of open superstring six-point amplitudes. Furthermore, we provide the low-energy expansion (in the inverse string tension) of our amplitude and discuss some relevant Dp-brane couplings associated to it. Finally, we write down an expression for the general structure of the disk amplitude of any number n_c of closed strings in terms of pure open string amplitudes involving 2n_c open strings., Comment: 64 pages, 11 figures

We classify thick irreducible 3-spherical twin buildings of rank at least 3 in which every panel contains at least 6 chambers. Together with the Main result of [11] we obtain a classification of thick irreducible 3-spherical twin buildings.