Single-molecule localization microscopy techniques, like stochastic optical reconstruction microscopy (STORM), visualize biological specimens by stochastically exciting sparse blinking emitters. The raw images suffer from unwanted background fluorescence, which must be removed to achieve super-resolution. We introduce a sparsity-based background removal method by adapting a neural network (SLNet) from a different microscopy domain. The SLNet computes a low-rank representation of the images, and then, by subtracting it from the raw images, the sparse component is computed, representing the frames without the background. We compared our approach with widely used background removal methods, such as the median background removal or the rolling ball algorithm, on two commonly used STORM datasets, one glial cell, and one microtubule dataset. The SLNet delivers STORM frames with less background, leading to higher emitters' localization precision and higher-resolution reconstructed images than commonly used methods. Notably, the SLNet is lightweight and easily trainable (<5 min). Since it is trained in an unsupervised manner, no prior information is required and can be applied to any STORM dataset. We uploaded a pre-trained SLNet to the Bioimage model zoo, easily accessible through ImageJ. Our results show that our sparse decomposition method could be an essential and efficient STORM pre-processing tool.
In this paper, we investigate a cascaded channel estimation method for a millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) system aided by a reconfigurable intelligent surface (RIS) with the BS equipped with low-resolution analog-to-digital converters (ADCs), where the BS and the RIS are both equipped with a uniform planar array (UPA). Due to the sparse property of mmWave channel, the channel estimation can be solved as a compressed sensing (CS) problem. However, the low-resolution quantization cause severe information loss of signals, and traditional CS algorithms are unable to work well. To recovery the signal and the sparse angular domain channel from quantization, we introduce Bayesian inference and efficient vector approximate message passing (VAMP) algorithm to solve the quantize output CS problem. To further improve the efficiency of the VAMP algorithm, a Fast Fourier Transform (FFT) based fast computation method is derived. Simulation results demonstrate the effectiveness and the accuracy of the proposed cascaded channel estimation method for the RIS-aided mmWave massive MIMO system with few-bit ADCs. Furthermore, the proposed channel estimation method can reach an acceptable performance gap between the low-resolution ADCs and the infinite ADCs for the low signal-to-noise ratio (SNR), which implies the applicability of few-bit ADCs in practice.
Accurate mapping of greenhouse gas fluxes at the Earth's surface is essential for the validation and calibration of climate models. In this study, we present a framework for surface flux estimation with drones. Our approach uses data assimilation (DA) to infer fluxes from drone-based observations, and reinforcement learning (RL) to optimize the drone's sampling strategy. Herein, we demonstrate that a RL-trained drone can quantify a CO2 hotspot more accurately than a drone sampling along a predefined flight path that traverses the emission plume. We find that information-based reward functions can match the performance of an error-based reward function that quantifies the difference between the estimated surface flux and the true value. Reward functions based on information gain and information entropy can motivate actions that increase the drone's confidence in its updated belief, without requiring knowledge of the true surface flux. These findings provide valuable insights for further development of the framework for the mapping of more complex surface flux fields.
Although few-shot action recognition based on metric learning paradigm has achieved significant success, it fails to address the following issues: (1) inadequate action relation modeling and underutilization of multi-modal information; (2) challenges in handling video matching problems with different lengths and speeds, and video matching problems with misalignment of video sub-actions. To address these issues, we propose a Two-Stream Joint Matching method based on contrastive learning (TSJM), which consists of two modules: Multi-modal Contrastive Learning Module (MCL) and Joint Matching Module (JMM). The objective of the MCL is to extensively investigate the inter-modal mutual information relationships, thereby thoroughly extracting modal information to enhance the modeling of action relationships. The JMM aims to simultaneously address the aforementioned video matching problems. The effectiveness of the proposed method is evaluated on two widely used few shot action recognition datasets, namely, SSv2 and Kinetics. Comprehensive ablation experiments are also conducted to substantiate the efficacy of our proposed approach.
This paper studies four Graph Neural Network architectures (GNNs) for a graph classification task on a synthetic dataset created using classic generative models of Network Science. Since the synthetic networks do not contain (node or edge) features, five different augmentation strategies (artificial feature types) are applied to nodes. All combinations of the 4 GNNs (GCN with Hierarchical and Global aggregation, GIN and GATv2) and the 5 feature types (constant 1, noise, degree, normalized degree and ID -- a vector of the number of cycles of various lengths) are studied and their performances compared as a function of the hidden dimension of artificial neural networks used in the GNNs. The generalisation ability of these models is also analysed using a second synthetic network dataset (containing networks of different sizes).Our results point towards the balanced importance of the computational power of the GNN architecture and the the information level provided by the artificial features. GNN architectures with higher computational power, like GIN and GATv2, perform well for most augmentation strategies. On the other hand, artificial features with higher information content, like ID or degree, not only consistently outperform other augmentation strategies, but can also help GNN architectures with lower computational power to achieve good performance.
Precise detection of tiny objects in remote sensing imagery remains a significant challenge due to their limited visual information and frequent occurrence within scenes. This challenge is further exacerbated by the practical burden and inherent errors associated with manual annotation: annotating tiny objects is laborious and prone to errors (i.e., label noise). Training detectors for such objects using noisy labels often leads to suboptimal performance, with networks tending to overfit on noisy labels. In this study, we address the intricate issue of tiny object detection under noisy label supervision. We systematically investigate the impact of various types of noise on network training, revealing the vulnerability of object detectors to class shifts and inaccurate bounding boxes for tiny objects. To mitigate these challenges, we propose a DeNoising Tiny Object Detector (DN-TOD), which incorporates a Class-aware Label Correction (CLC) scheme to address class shifts and a Trend-guided Learning Strategy (TLS) to handle bounding box noise. CLC mitigates inaccurate class supervision by identifying and filtering out class-shifted positive samples, while TLS reduces noisy box-induced erroneous supervision through sample reweighting and bounding box regeneration. Additionally, Our method can be seamlessly integrated into both one-stage and two-stage object detection pipelines. Comprehensive experiments conducted on synthetic (i.e., noisy AI-TOD-v2.0 and DOTA-v2.0) and real-world (i.e., AI-TOD) noisy datasets demonstrate the robustness of DN-TOD under various types of label noise. Notably, when applied to the strong baseline RFLA, DN-TOD exhibits a noteworthy performance improvement of 4.9 points under 40% mixed noise. Datasets, codes, and models will be made publicly available.
Sentiment analysis and emotion detection are important research topics in natural language processing (NLP) and benefit many downstream tasks. With the widespread application of LLMs, researchers have started exploring the application of LLMs based on instruction-tuning in the field of sentiment analysis. However, these models only focus on single aspects of affective classification tasks (e.g. sentimental polarity or categorical emotions), and overlook the regression tasks (e.g. sentiment strength or emotion intensity), which leads to poor performance in downstream tasks. The main reason is the lack of comprehensive affective instruction tuning datasets and evaluation benchmarks, which cover various affective classification and regression tasks. Moreover, although emotional information is useful for downstream tasks, existing downstream datasets lack high-quality and comprehensive affective annotations. In this paper, we propose EmoLLMs, the first series of open-sourced instruction-following LLMs for comprehensive affective analysis based on fine-tuning various LLMs with instruction data, the first multi-task affective analysis instruction dataset (AAID) with 234K data samples based on various classification and regression tasks to support LLM instruction tuning, and a comprehensive affective evaluation benchmark (AEB) with 14 tasks from various sources and domains to test the generalization ability of LLMs. We propose a series of EmoLLMs by fine-tuning LLMs with AAID to solve various affective instruction tasks. We compare our model with a variety of LLMs on AEB, where our models outperform all other open-sourced LLMs, and surpass ChatGPT and GPT-4 in most tasks, which shows that the series of EmoLLMs achieve the ChatGPT-level and GPT-4-level generalization capabilities on affective analysis tasks, and demonstrates our models can be used as affective annotation tools.
Infrared and visible image fusion aims at generating a fused image containing the intensity and detail information of source images, and the key issue is effectively measuring and integrating the complementary information of multi-modality images from the same scene. Existing methods mostly adopt a simple weight in the loss function to decide the information retention of each modality rather than adaptively measuring complementary information for different image pairs. In this study, we propose a multi-scale dual attention (MDA) framework for infrared and visible image fusion, which is designed to measure and integrate complementary information in both structure and loss function at the image and patch level. In our method, the residual downsample block decomposes source images into three scales first. Then, dual attention fusion block integrates complementary information and generates a spatial and channel attention map at each scale for feature fusion. Finally, the output image is reconstructed by the residual reconstruction block. Loss function consists of image-level, feature-level and patch-level three parts, of which the calculation of the image-level and patch-level two parts are based on the weights generated by the complementary information measurement. Indeed, to constrain the pixel intensity distribution between the output and infrared image, a style loss is added. Our fusion results perform robust and informative across different scenarios. Qualitative and quantitative results on two datasets illustrate that our method is able to preserve both thermal radiation and detailed information from two modalities and achieve comparable results compared with the other state-of-the-art methods. Ablation experiments show the effectiveness of our information integration architecture and adaptively measure complementary information retention in the loss function.
Neural solvers based on attention mechanism have demonstrated remarkable effectiveness in solving vehicle routing problems. However, in the generalization process from small scale to large scale, we find a phenomenon of the dispersion of attention scores in existing neural solvers, which leads to poor performance. To address this issue, this paper proposes a distance-aware attention reshaping method, assisting neural solvers in solving large-scale vehicle routing problems. Specifically, without the need for additional training, we utilize the Euclidean distance information between current nodes to adjust attention scores. This enables a neural solver trained on small-scale instances to make rational choices when solving a large-scale problem. Experimental results show that the proposed method significantly outperforms existing state-of-the-art neural solvers on the large-scale CVRPLib dataset.
Creating and deploying customized applications is crucial for operational success and enriching user experiences in the rapidly evolving modern business world. A prominent facet of modern user experiences is the integration of chatbots or voice assistants. The rapid evolution of Large Language Models (LLMs) has provided a powerful tool to build conversational applications. We present Walert, a customized LLM-based conversational agent able to answer frequently asked questions about computer science degrees and programs at RMIT University. Our demo aims to showcase how conversational information-seeking researchers can effectively communicate the benefits of using best practices to stakeholders interested in developing and deploying LLM-based chatbots. These practices are well-known in our community but often overlooked by practitioners who may not have access to this knowledge. The methodology and resources used in this demo serve as a bridge to facilitate knowledge transfer from experts, address industry professionals' practical needs, and foster a collaborative environment. The data and code of the demo are available at https://github.com/rmit-ir/walert.