GNN-LoFI: a Novel Graph Neural Network through Localized Feature-based Histogram Intersection

Graph neural networks are increasingly becoming the framework of choice for graph-based machine learning. In this paper, we propose a new graph neural network architecture that substitutes classical message passing with an analysis of the local distribution of node features. To this end, we extract the distribution of features in the egonet for each local neighbourhood and compare them against a set of learned label distributions by taking the histogram intersection kernel. The similarity information is then propagated to other nodes in the network, effectively creating a message passing-like mechanism where the message is determined by the ensemble of the features. We perform an ablation study to evaluate the network's performance under different choices of its hyper-parameters. Finally, we test our model on standard graph classification and regression benchmarks, and we find that it outperforms widely used alternative approaches, including both graph kernels and graph neural networks.

DK-SLAM: Monocular Visual SLAM with Deep Keypoints Adaptive Learning, Tracking and Loop-Closing

Unreliable feature extraction and matching in handcrafted features undermine the performance of visual SLAM in complex real-world scenarios. While learned local features, leveraging CNNs, demonstrate proficiency in capturing high-level information and excel in matching benchmarks, they encounter challenges in continuous motion scenes, resulting in poor generalization and impacting loop detection accuracy. To address these issues, we present DK-SLAM, a monocular visual SLAM system with adaptive deep local features. MAML optimizes the training of these features, and we introduce a coarse-to-fine feature tracking approach. Initially, a direct method approximates the relative pose between consecutive frames, followed by a feature matching method for refined pose estimation. To counter cumulative positioning errors, a novel online learning binary feature-based online loop closure module identifies loop nodes within a sequence. Experimental results underscore DK-SLAM's efficacy, outperforms representative SLAM solutions, such as ORB-SLAM3 on publicly available datasets.

The Rise of Diffusion Models in Time-Series Forecasting

This survey delves into the application of diffusion models in time-series forecasting. Diffusion models are demonstrating state-of-the-art results in various fields of generative AI. The paper includes comprehensive background information on diffusion models, detailing their conditioning methods and reviewing their use in time-series forecasting. The analysis covers 11 specific time-series implementations, the intuition and theory behind them, the effectiveness on different datasets, and a comparison among each other. Key contributions of this work are the thorough exploration of diffusion models' applications in time-series forecasting and a chronologically ordered overview of these models. Additionally, the paper offers an insightful discussion on the current state-of-the-art in this domain and outlines potential future research directions. This serves as a valuable resource for researchers in AI and time-series analysis, offering a clear view of the latest advancements and future potential of diffusion models.

Advancing Large Multi-modal Models with Explicit Chain-of-Reasoning and Visual Question Generation

The increasing demand for intelligent systems capable of interpreting and reasoning about visual content requires the development of Large Multi-Modal Models (LMMs) that are not only accurate but also have explicit reasoning capabilities. This paper presents a novel approach to imbue an LMM with the ability to conduct explicit reasoning based on visual content and textual instructions. We introduce a system that can ask a question to acquire necessary knowledge, thereby enhancing the robustness and explicability of the reasoning process. Our method comprises the development of a novel dataset generated by a Large Language Model (LLM), designed to promote chain-of-thought reasoning combined with a question-asking mechanism. We designed an LMM, which has high capabilities on region awareness to address the intricate requirements of image-text alignment. The model undergoes a three-stage training phase, starting with large-scale image-text alignment using a large-scale datasets, followed by instruction tuning, and fine-tuning with a focus on chain-of-thought reasoning. The results demonstrate a stride toward a more robust, accurate, and interpretable LMM, capable of reasoning explicitly and seeking information proactively when confronted with ambiguous visual input.

Measuring Object Rotation via Visuo-Tactile Segmentation

When carrying out robotic manipulation tasks, objects occasionally fall as a result of the rotation caused by slippage. This can be prevented by obtaining tactile information that provides better knowledge on the physical properties of the grasping. In this paper, we estimate the rotation angle of a grasped object when slippage occurs. We implement a system made up of a neural network with which to segment the contact region and an algorithm with which to estimate the rotated angle of that region. This method is applied to DIGIT tactile sensors. Our system has additionally been trained and tested with our publicly available dataset which is, to the best of our knowledge, the first dataset related to tactile segmentation from non-synthetic images to appear in the literature, and with which we have attained results of 95% and 90% as regards Dice and IoU metrics in the worst scenario. Moreover, we have obtained a maximum error of 3 degrees when testing with objects not previously seen by our system in 45 different lifts. This, therefore, proved that our approach is able to detect the slippage movement, thus providing a possible reaction that will prevent the object from falling.

Temporal Insight Enhancement: Mitigating Temporal Hallucination in Multimodal Large Language Models

Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced the comprehension of multimedia content, bringing together diverse modalities such as text, images, and videos. However, a critical challenge faced by these models, especially when processing video inputs, is the occurrence of hallucinations - erroneous perceptions or interpretations, particularly at the event level. This study introduces an innovative method to address event-level hallucinations in MLLMs, focusing on specific temporal understanding in video content. Our approach leverages a novel framework that extracts and utilizes event-specific information from both the event query and the provided video to refine MLLMs' response. We propose a unique mechanism that decomposes on-demand event queries into iconic actions. Subsequently, we employ models like CLIP and BLIP2 to predict specific timestamps for event occurrences. Our evaluation, conducted using the Charades-STA dataset, demonstrates a significant reduction in temporal hallucinations and an improvement in the quality of event-related responses. This research not only provides a new perspective in addressing a critical limitation of MLLMs but also contributes a quantitatively measurable method for evaluating MLLMs in the context of temporal-related questions.

Learned Image resizing with efficient training (LRET) facilitates improved performance of large-scale digital histopathology image classification models

Histologic examination plays a crucial role in oncology research and diagnostics. The adoption of digital scanning of whole slide images (WSI) has created an opportunity to leverage deep learning-based image classification methods to enhance diagnosis and risk stratification. Technical limitations of current approaches to training deep convolutional neural networks (DCNN) result in suboptimal model performance and make training and deployment of comprehensive classification models unobtainable. In this study, we introduce a novel approach that addresses the main limitations of traditional histopathology classification model training. Our method, termed Learned Resizing with Efficient Training (LRET), couples efficient training techniques with image resizing to facilitate seamless integration of larger histology image patches into state-of-the-art classification models while preserving important structural information. We used the LRET method coupled with two distinct resizing techniques to train three diverse histology image datasets using multiple diverse DCNN architectures. Our findings demonstrate a significant enhancement in classification performance and training efficiency. Across the spectrum of experiments, LRET consistently outperforms existing methods, yielding a substantial improvement of 15-28% in accuracy for a large-scale, multiclass tumor classification task consisting of 74 distinct brain tumor types. LRET not only elevates classification accuracy but also substantially reduces training times, unlocking the potential for faster model development and iteration. The implications of this work extend to broader applications within medical imaging and beyond, where efficient integration of high-resolution images into deep learning pipelines is paramount for driving advancements in research and clinical practice.

Bounding Consideration Probabilities in Consider-Then-Choose Ranking Models

A common theory of choice posits that individuals make choices in a two-step process, first selecting some subset of the alternatives to consider before making a selection from the resulting consideration set. However, inferring unobserved consideration sets (or item consideration probabilities) in this "consider then choose" setting poses significant challenges, because even simple models of consideration with strong independence assumptions are not identifiable, even if item utilities are known. We consider a natural extension of consider-then-choose models to a top-$k$ ranking setting, where we assume rankings are constructed according to a Plackett-Luce model after sampling a consideration set. While item consideration probabilities remain non-identified in this setting, we prove that knowledge of item utilities allows us to infer bounds on the relative sizes of consideration probabilities. Additionally, given a condition on the expected consideration set size, we derive absolute upper and lower bounds on item consideration probabilities. We also provide algorithms to tighten those bounds on consideration probabilities by propagating inferred constraints. Thus, we show that we can learn useful information about consideration probabilities despite not being able to identify them precisely. We demonstrate our methods on a ranking dataset from a psychology experiment with two different ranking tasks (one with fixed consideration sets and one with unknown consideration sets). This combination of data allows us to estimate utilities and then learn about unknown consideration probabilities using our bounds.

On the selection of the correct number of terms for profile construction: theoretical and empirical analysis

In this paper, we examine the problem of building a user profile from a set of documents. This profile will consist of a subset of the most representative terms in the documents that best represent user preferences or interests. Inspired by the discrete concentration theory we have conducted an axiomatic study of seven properties that a selection function should fulfill: the minimum and maximum uncertainty principle, invariant to adding zeros, invariant to scale transformations, principle of nominal increase, transfer principle and the richest get richer inequality. We also present a novel selection function based on the use of similarity metrics, and more specifically the cosine measure which is commonly used in information retrieval, and demonstrate that this verifies six of the properties in addition to a weaker variant of the transfer principle, thereby representing a good selection approach. The theoretical study was complemented with an empirical study to compare the performance of different selection criteria (weight- and unweight-based) using real data in a parliamentary setting. In this study, we analyze the performance of the different functions focusing on the two main factors affecting the selection process: profile size (number of terms) and weight distribution. These profiles are then used in a document filtering task to show that our similarity-based approach performs well in terms not only of recommendation accuracy but also efficiency (we obtain smaller profiles and consequently faster recommendations).

A Framework for Agricultural Food Supply Chain using Blockchain

The main aim of the paper is to create a trust and transparency in the food supply chain system, ensuring food safety for everyone with the help of Blockchain Technology. Food supply chain is the process of tracing a crop from the farmer or producer to the buyer. With the advent of blockchain, providing a safe and fraud-free environment for the provision of numerous agricultural necessities has become much easier. Because of the globalization of trade, the present supply chain market today includes various companies involving integration of data, complex transactions and distribution. Information tamper resistance, supply-demand relationships, and traceable oversight are all difficulties that arise as a result of this. Blockchain is a distributed ledger technology that can provide information that is resistant to tampering. This strategy can eliminate the need for a centralized trusted authority, intermediaries, and business histories, allowing for increased production and security while maintaining the highest levels of integrity, liability, and safety. In order to have an integrity and transparency in food supply chain in the agricultural sector, a framework is proposed here based on block chain and IoT.