Search

Fluorescence microscopes can record the dynamics of living cells with high spatio-temporal resolution in a single plane. However, monitoring rapid and dim fluorescence fluctuations, e.g induced by neuronal activity in the brain, remains challenging for 3D-distributed emitters due to out-of-focus fluorescence background, a restricted photon budget, and the speed limit of conventional scanning systems. Here, we introduce a Thermally Adaptive Surface strategy, capable of simultaneously recording, at camera framerate, the activity of 3D-distributed objects. This innovative microscope leverages on an array of thermally tuneable microlenses that offer low chromatic aberration and high transmission, and can be combined with patterned illumination to provide optical sectioning. We demonstrate its potential in vivo, by simultaneously monitoring fast fluorescent dynamics at different depths in the zebrafish larval brain, at a rate of 0.5 kHz and over a large field of view (360um x 360um).

Multiple Instance Learning (MIL) methods allow for gigapixel Whole-Slide Image (WSI) analysis with only slide-level annotations. Interpretability is crucial for safely deploying such algorithms in high-stakes medical domains. Traditional MIL methods offer explanations by highlighting salient regions. However, such spatial heatmaps provide limited insights for end users. To address this, we propose a novel inherently interpretable WSI-classification approach that uses human-understandable pathology concepts to generate explanations. Our proposed Concept MIL model leverages recent advances in vision-language models to directly predict pathology concepts based on image features. The model's predictions are obtained through a linear combination of the concepts identified on the top-K patches of a WSI, enabling inherent explanations by tracing each concept's influence on the prediction. In contrast to traditional concept-based interpretable models, our approach eliminates the need for costly human annotations by leveraging the vision-language model. We validate our method on two widely used pathology datasets: Camelyon16 and PANDA. On both datasets, Concept MIL achieves AUC and accuracy scores over 0.9, putting it on par with state-of-the-art models. We further find that 87.1\% (Camelyon16) and 85.3\% (PANDA) of the top 20 patches fall within the tumor region. A user study shows that the concepts identified by our model align with the concepts used by pathologists, making it a promising strategy for human-interpretable WSI classification.

Motivation: Students learning to program often reach states where they are stuck and can make no forward progress. An automatically generated next-step hint can help them make forward progress and support their learning. It is important to know what makes a good hint or a bad hint, and how to generate good hints automatically in novice programming tools, for example using Large Language Models (LLMs). Method and participants: We recruited 44 Java educators from around the world to participate in an online study. We used a set of real student code states as hint-generation scenarios. Participants used a technique known as comparative judgement to rank a set of candidate next-step Java hints, which were generated by Large Language Models (LLMs) and by five human experienced educators. Participants ranked the hints without being told how they were generated. Findings: We found that LLMs had considerable variation in generating high quality next-step hints for programming novices, with GPT-4 outperforming other models tested. When used with a well-designed prompt, GPT-4 outperformed human experts in generating pedagogically valuable hints. A multi-stage prompt was the most effective LLM prompt. We found that the two most important factors of a good hint were length (80--160 words being best), and reading level (US grade 9 or below being best). Offering alternative approaches to solving the problem was considered bad, and we found no effect of sentiment. Conclusions: Automatic generation of these hints is immediately viable, given that LLMs outperformed humans -- even when the students' task is unknown. The fact that only the best prompts achieve this outcome suggests that students on their own are unlikely to be able to produce the same benefit. The prompting task, therefore, should be embedded in an expert-designed tool., Comment: 36 pages, 14 figures, 7 tables

Many processes in microfluidics and biology are driven or affected by viscosity. While several methods are able to measure this parameter globally, very few can provide high resolution viscosity images. Optimizing the locality of viscosity measurements demands smaller probes but also shorter lateral diffusion lengths and measurement times. Here, we propose to use sub-micrometer magnetic rods to perform high resolution viscosity imaging. An external magnetic field forces the oscillation of superparamagnetic iron oxide rods. Under linearly polarized illumination, the rotation of these highly anisotropic optical scatterers induces a blinking which is analyzed by heterodyne holography. The spectral analysis of the rotation dynamics yields a regime transition frequency from which the local viscosity is deduced. Holography provides a 3D optical field reconstruction and 3D superlocalization of the rods, which allows super-resolved viscosity measurements. Relying on the fast Brownian rotation instead of the slower translation component of nanorods therefore allows faster measurements and, crucially, smaller effective voxels for viscosity determination. We thus demonstrate that viscosity imaging is possible with a 0.5 (micron)^3 3D-resolution., Comment: 14 pages, 4 figures, 49 references

Liver tumor ablation procedures require accurate placement of the needle applicator at the tumor centroid. The lower-cost and real-time nature of ultrasound (US) has advantages over computed tomography (CT) for applicator guidance, however, in some patients, liver tumors may be occult on US and tumor mimics can make lesion identification challenging. Image registration techniques can aid in interpreting anatomical details and identifying tumors, but their clinical application has been hindered by the tradeoff between alignment accuracy and runtime performance, particularly when compensating for liver motion due to patient breathing or movement. Therefore, we propose a 2D-3D US registration approach to enable intra-procedural alignment that mitigates errors caused by liver motion. Specifically, our approach can correlate imbalanced 2D and 3D US image features and use continuous 6D rotation representations to enhance the model's training stability. The dataset was divided into 2388, 196 and 193 image pairs for training, validation and testing, respectively. Our approach achieved a mean Euclidean distance error of 2.28 mm $\pm$ 1.81 mm and a mean geodesic angular error of 2.99$^{\circ}$ $\pm$ 1.95$^{\circ}$, with a runtime of 0.22 seconds per 2D-3D US image pair. These results demonstrate that our approach can achieve accurate alignment and clinically acceptable runtime, indicating potential for clinical translation., Comment: 15 pagers, 9 figures

In automotive engineering, designing for optimal vehicle dynamics is challenging due to the complexities involved in analysing the behaviour of a multibody system. Typically, a simplified set of dynamics equations for only the key bodies of the vehicle such as the chassis and wheels are formulated while reducing their degrees of freedom. In contrast, one could employ high-fidelity multibody dynamics simulation and include more intricate details such as the individual suspension components while considering full degrees of freedom for all bodies; however, this is more computationally demanding. Also, for gradient-based design optimization, computing adjoints for different objective functions can be more challenging for the latter approach, and often not feasible if an existing multibody dynamics solver is used. We propose a mixed-fidelity multidisciplinary approach, in which a simplified set of dynamics equations are used to model the whole vehicle while incorporating a high-fidelity multibody suspension module as an additional coupled discipline. We then employ MAUD (modular analysis and unified derivatives) to combine analytical derivatives based on the dynamics equations and finite differences obtained using an existing multibody solver. Also, we use a collocation method for time integration, which solves for both the system trajectory and optimal design variables simultaneously. The benefits of our approach are shown in an experiment conducted to find optimal vehicle parameters that optimize ride comfort and driving performance considering vertical vehicle dynamics., Comment: ECCOMAS Congress 2024

The integration of Field Programmable Gate Arrays (FPGAs) into cloud computing systems has become commonplace. As the operating systems used to manage these systems evolve, special consideration must be given to DRAM devices accessible by FPGAs. These devices may hold sensitive data that can become inadvertently exposed to adversaries following user logout. Although addressed in some cloud FPGA environments, automatic DRAM clearing after process termination is not automatically included in popular FPGA runtime environments nor in most proposed cloud FPGA hypervisors. In this paper, we examine DRAM data persistence in AMD/Xilinx Alveo U280 nodes that are part of the Open Cloud Testbed (OCT). Our results indicate that DDR4 DRAM is not automatically cleared following user logout from an allocated node and subsequent node users can easily obtain recognizable data from the DRAM following node reallocation over 17 minutes later. This issue is particularly relevant for systems which support FPGA multi-tenancy.

Self-supervised learning (SSL) representations from massively multilingual models offer a promising solution for low-resource language speech tasks. Despite advancements, language adaptation in TTS systems remains an open problem. This paper explores the language adaptation capability of ZMM-TTS, a recent SSL-based multilingual TTS system proposed in our previous work. We conducted experiments on 12 languages using limited data with various fine-tuning configurations. We demonstrate that the similarity in phonetics between the pre-training and target languages, as well as the language category, affects the target language's adaptation performance. Additionally, we find that the fine-tuning dataset size and number of speakers influence adaptability. Surprisingly, we also observed that using paired data for fine-tuning is not always optimal compared to audio-only data. Beyond speech intelligibility, our analysis covers speaker similarity, language identification, and predicted MOS., Comment: Accepted to Interspeech 2024

Requirements elicitation, a critical, yet time-consuming and challenging step in product development, often fails to capture the full spectrum of user needs. This may lead to products that fall short of expectations. This paper introduces a novel framework that leverages Large Language Models (LLMs) to automate and enhance the requirements elicitation process. LLMs are used to generate a vast array of simulated users (LLM agents), enabling the exploration of a much broader range of user needs and unforeseen use cases. These agents engage in product experience scenarios, through explaining their actions, observations, and challenges. Subsequent agent interviews and analysis uncover valuable user needs, including latent ones. We validate our framework with three experiments. First, we explore different methodologies for diverse agent generation, discussing their advantages and shortcomings. We measure the diversity of identified user needs and demonstrate that context-aware agent generation leads to greater diversity. Second, we show how our framework effectively mimics empathic lead user interviews, identifying a greater number of latent needs than conventional human interviews. Third, we showcase that LLMs can be used to analyze interviews, capture needs, and classify them as latent or not. Our work highlights the potential of using LLM agents to accelerate early-stage product development, reduce costs, and increase innovation.

As mechanical systems become more complex and technological advances accelerate, the traditional reliance on heritage designs for engineering endeavors is being diminished in its effectiveness. Considering the dynamic nature of the design industry where new challenges are continually emerging, alternative sources of knowledge need to be sought to guide future design efforts. One promising avenue lies in the analysis of design optimization data, which has the potential to offer valuable insights and overcome the limitations of heritage designs. This paper presents a step toward extracting knowledge from optimization data in multi-split fluid-based thermal management systems using different classification machine learning methods, so that designers can use it to guide decisions in future design efforts. This approach offers several advantages over traditional design heritage methods, including applicability in cases where there is no design heritage and the ability to derive optimal designs. We showcase our framework through four case studies with varying levels of complexity. These studies demonstrate its effectiveness in enhancing the design of complex thermal management systems. Our results show that the knowledge extracted from the configuration design optimization data provides a good basis for more general design of complex thermal management systems. It is shown that the objective value of the estimated optimal configuration closely approximates the true optimal configuration with less than 1 percent error, achieved using basic features based on the system heat loads without involving the corresponding optimal open loop control (OLOC) features. This eliminates the need to solve the OLOC problem, leading to reduced computation costs., Comment: 13 pages, 20 figures

High power density systems require efficient cooling to maintain their thermal performance. Despite this, as systems get larger and more complex, human practice and insight may not suffice to determine the desired thermal management system designs. To this end, a framework for automatic architecture exploration is presented in this article for a class of single-phase, multi-split cooling systems. For this class of systems, heat generation devices are clustered based on their spatial information, and flow-split are added only when required and at the location of heat devices. To generate different architectures, candidate architectures are represented as graphs. From these graphs, dynamic physics models are created automatically using a graph-based thermal modeling framework. Then, an optimal fluid flow distribution problem is solved by addressing temperature constraints in the presence of exogenous heat loads to achieve optimal performance. The focus in this work is on the design of general multi-split heat management systems. The architectures discussed here can be used for various applications in the domain of configuration design. The multi-split algorithm can produce configurations where splitting can occur at any of the vertices. The results presented include 3 categories of cases and are discussed in detail., Comment: 11 pages, 18 figures

Pruning is a compression method which aims to improve the efficiency of neural networks by reducing their number of parameters while maintaining a good performance, thus enhancing the performance-to-cost ratio in nontrivial ways. Of particular interest are structured pruning techniques, in which whole portions of parameters are removed altogether, resulting in easier to leverage shrunk architectures. Since its growth in popularity in the recent years, pruning gave birth to countless papers and contributions, resulting first in critical inconsistencies in the way results are compared, and then to a collective effort to establish standardized benchmarks. However, said benchmarks are based on training practices that date from several years ago and do not align with current practices. In this work, we verify how results in the recent literature of pruning hold up against networks that underwent both state-of-the-art training methods and trivial model scaling. We find that the latter clearly and utterly outperform all the literature we compared to, proving that updating standard pruning benchmarks and re-evaluating classical methods in their light is an absolute necessity. We thus introduce a new challenging baseline to compare structured pruning to: ThinResNet., Comment: 11 pages, 2 figures, 3 tables