JMLR: Joint Medical LLM and Retrieval Training for Enhancing Reasoning and Professional Question Answering Capability

With the explosive growth of medical data and the rapid development of artificial intelligence technology, precision medicine has emerged as a key to enhancing the quality and efficiency of healthcare services. In this context, Large Language Models (LLMs) play an increasingly vital role in medical knowledge acquisition and question-answering systems. To further improve the performance of these systems in the medical domain, we introduce an innovative method that jointly trains an Information Retrieval (IR) system and an LLM during the fine-tuning phase. This approach, which we call Joint Medical LLM and Retrieval Training (JMLR), is designed to overcome the challenges faced by traditional models in handling medical question-answering tasks. By employing a synchronized training mechanism, JMLR reduces the demand for computational resources and enhances the model's ability to leverage medical knowledge for reasoning and answering questions. Our experimental results demonstrate that JMLR-13B (81.2% on Amboos, 61.3% on MedQA) outperforms models using conventional pre-training and fine-tuning Meditron-70B (76.4% on AMBOSS, 60.3% on MedQA). For models of the same 7B scale, JMLR-7B(68.7% on Amboos, 51.7% on MedQA) significantly outperforms other public models (Meditron-7B: 50.1%, 47.9%), proving its superiority in terms of cost (our training time: 37 hours, traditional method: 144 hours), efficiency, and effectiveness in medical question-answering tasks. Through this work, we provide a new and efficient knowledge enhancement tool for healthcare, demonstrating the great potential of integrating IR and LLM training in precision medical information retrieval and question-answering systems.

CharNeRF: 3D Character Generation from Concept Art

3D modeling holds significant importance in the realms of AR/VR and gaming, allowing for both artistic creativity and practical applications. However, the process is often time-consuming and demands a high level of skill. In this paper, we present a novel approach to create volumetric representations of 3D characters from consistent turnaround concept art, which serves as the standard input in the 3D modeling industry. While Neural Radiance Field (NeRF) has been a game-changer in image-based 3D reconstruction, to the best of our knowledge, there is no known research that optimizes the pipeline for concept art. To harness the potential of concept art, with its defined body poses and specific view angles, we propose encoding it as priors for our model. We train the network to make use of these priors for various 3D points through a learnable view-direction-attended multi-head self-attention layer. Additionally, we demonstrate that a combination of ray sampling and surface sampling enhances the inference capabilities of our network. Our model is able to generate high-quality 360-degree views of characters. Subsequently, we provide a simple guideline to better leverage our model to extract the 3D mesh. It is important to note that our model's inferencing capabilities are influenced by the training data's characteristics, primarily focusing on characters with a single head, two arms, and two legs. Nevertheless, our methodology remains versatile and adaptable to concept art from diverse subject matters, without imposing any specific assumptions on the data.

An Empirical Study of the Generalization Ability of Lidar 3D Object Detectors to Unseen Domains

3D Object Detectors (3D-OD) are crucial for understanding the environment in many robotic tasks, especially autonomous driving. Including 3D information via Lidar sensors improves accuracy greatly. However, such detectors perform poorly on domains they were not trained on, i.e. different locations, sensors, weather, etc., limiting their reliability in safety-critical applications. There exist methods to adapt 3D-ODs to these domains; however, these methods treat 3D-ODs as a black box, neglecting underlying architectural decisions and source-domain training strategies. Instead, we dive deep into the details of 3D-ODs, focusing our efforts on fundamental factors that influence robustness prior to domain adaptation. We systematically investigate four design choices (and the interplay between them) often overlooked in 3D-OD robustness and domain adaptation: architecture, voxel encoding, data augmentations, and anchor strategies. We assess their impact on the robustness of nine state-of-the-art 3D-ODs across six benchmarks encompassing three types of domain gaps - sensor type, weather, and location. Our main findings are: (1) transformer backbones with local point features are more robust than 3D CNNs, (2) test-time anchor size adjustment is crucial for adaptation across geographical locations, significantly boosting scores without retraining, (3) source-domain augmentations allow the model to generalize to low-resolution sensors, and (4) surprisingly, robustness to bad weather is improved when training directly on more clean weather data than on training with bad weather data. We outline our main conclusions and findings to provide practical guidance on developing more robust 3D-ODs.

Movable Antenna-Enabled Co-Frequency Co-Time Full-Duplex Wireless Communication

Movable antenna (MA) provides an innovative way to arrange antennas that can contribute to improved signal quality and more effective interference management. This method is especially beneficial for co-frequency co-time full-duplex (CCFD) wireless communication, which struggles with self-interference (SI) that usually overpowers the desired incoming signals. By dynamically repositioning transmit/receive antennas, we can mitigate the SI and enhance the reception of incoming signals. Thus, this paper proposes a novel MA-enabled point-to-point CCFD system and formulates the minimum achievable rate of two CCFD terminals. To maximize the minimum achievable rate and determine the near-optimal positions of the MAs, we introduce a solution based on projected particle swarm optimization (PPSO), which can circumvent common suboptimal positioning issues. Moreover, numerical results reveal that the PPSO method leads to a better performance compared to the conventional alternating position optimization (APO). The results also demonstrate that an MA-enabled CCFD system outperforms the one using fixed-position antennas (FPAs).

MSTAR: Multi-Scale Backbone Architecture Search for Timeseries Classification

Most of the previous approaches to Time Series Classification (TSC) highlight the significance of receptive fields and frequencies while overlooking the time resolution. Hence, unavoidably suffered from scalability issues as they integrated an extensive range of receptive fields into classification models. Other methods, while having a better adaptation for large datasets, require manual design and yet not being able to reach the optimal architecture due to the uniqueness of each dataset. We overcome these challenges by proposing a novel multi-scale search space and a framework for Neural architecture search (NAS), which addresses both the problem of frequency and time resolution, discovering the suitable scale for a specific dataset. We further show that our model can serve as a backbone to employ a powerful Transformer module with both untrained and pre-trained weights. Our search space reaches the state-of-the-art performance on four datasets on four different domains while introducing more than ten highly fine-tuned models for each data.

Machine-learning prediction of tipping and collapse of the Atlantic Meridional Overturning Circulation

Recent research on the Atlantic Meridional Overturning Circulation (AMOC) raised concern about its potential collapse through a tipping point due to the climate-change caused increase in the freshwater input into the North Atlantic. The predicted time window of collapse is centered about the middle of the century and the earliest possible start is approximately two years from now. More generally, anticipating a tipping point at which the system transitions from one stable steady state to another is relevant to a broad range of fields. We develop a machine-learning approach to predicting tipping in noisy dynamical systems with a time-varying parameter and test it on a number of systems including the AMOC, ecological networks, an electrical power system, and a climate model. For the AMOC, our prediction based on simulated fingerprint data and real data of the sea surface temperature places the time window of a potential collapse between the years 2040 and 2065.

Smooth real-time motion planning based on a cascade dual-quaternion screw-geometry MPC

This paper investigates the tracking problem of a smooth coordinate-invariant trajectory using dual quaternion algebra. The proposed architecture consists of a cascade structure in which the outer-loop MPC performs real-time smoothing of the manipulator's end-effector twist while an inner-loop kinematic controller ensures tracking of the instantaneous desired end-effector pose. Experiments on a $7$-DoF Franka Emika Panda robotic manipulator validate the proposed method demonstrating its application to constraint the robot twists, accelerations and jerks within prescribed bounds.

Reading Relevant Feature from Global Representation Memory for Visual Object Tracking

Reference features from a template or historical frames are crucial for visual object tracking. Prior works utilize all features from a fixed template or memory for visual object tracking. However, due to the dynamic nature of videos, the required reference historical information for different search regions at different time steps is also inconsistent. Therefore, using all features in the template and memory can lead to redundancy and impair tracking performance. To alleviate this issue, we propose a novel tracking paradigm, consisting of a relevance attention mechanism and a global representation memory, which can adaptively assist the search region in selecting the most relevant historical information from reference features. Specifically, the proposed relevance attention mechanism in this work differs from previous approaches in that it can dynamically choose and build the optimal global representation memory for the current frame by accessing cross-frame information globally. Moreover, it can flexibly read the relevant historical information from the constructed memory to reduce redundancy and counteract the negative effects of harmful information. Extensive experiments validate the effectiveness of the proposed method, achieving competitive performance on five challenging datasets with 71 FPS.

Deep Rating Elicitation for New Users in Collaborative Filtering

Recent recommender systems started to use rating elicitation, which asks new users to rate a small seed itemset for inferring their preferences, to improve the quality of initial recommendations. The key challenge of the rating elicitation is to choose the seed items which can best infer the new users' preference. This paper proposes a novel end-to-end Deep learning framework for Rating Elicitation (DRE), that chooses all the seed items at a time with consideration of the non-linear interactions. To this end, it first defines categorical distributions to sample seed items from the entire itemset, then it trains both the categorical distributions and a neural reconstruction network to infer users' preferences on the remaining items from CF information of the sampled seed items. Through the end-to-end training, the categorical distributions are learned to select the most representative seed items while reflecting the complex non-linear interactions. Experimental results show that DRE outperforms the state-of-the-art approaches in the recommendation quality by accurately inferring the new users' preferences and its seed itemset better represents the latent space than the seed itemset obtained by the other methods.

On the Generalization Capability of Temporal Graph Learning Algorithms: Theoretical Insights and a Simpler Method

Temporal Graph Learning (TGL) has become a prevalent technique across diverse real-world applications, especially in domains where data can be represented as a graph and evolves over time. Although TGL has recently seen notable progress in algorithmic solutions, its theoretical foundations remain largely unexplored. This paper aims at bridging this gap by investigating the generalization ability of different TGL algorithms (e.g., GNN-based, RNN-based, and memory-based methods) under the finite-wide over-parameterized regime. We establish the connection between the generalization error of TGL algorithms and "the number of layers/steps" in the GNN-/RNN-based TGL methods and "the feature-label alignment (FLA) score", where FLA can be used as a proxy for the expressive power and explains the performance of memory-based methods. Guided by our theoretical analysis, we propose Simplified-Temporal-Graph-Network, which enjoys a small generalization error, improved overall performance, and lower model complexity. Extensive experiments on real-world datasets demonstrate the effectiveness of our method. Our theoretical findings and proposed algorithm offer essential insights into TGL from a theoretical standpoint, laying the groundwork for the designing practical TGL algorithms in future studies.