We present RelCon, a novel self-supervised Relative Contrastive learning approach for training a motion foundation model from wearable accelerometry sensors. First, a learnable distance measure is trained to capture motif similarity and domain-specific semantic information such as rotation invariance. Then, the learned distance provides a measurement of semantic similarity between a pair of accelerometry time-series, which we use to train our foundation model to model relative relationships across time and across subjects. The foundation model is trained on 1 billion segments from 87,376 participants, and achieves strong performance across multiple downstream tasks, including human activity recognition and gait metric regression. To our knowledge, we are the first to show the generalizability of a foundation model with motion data from wearables across distinct evaluation tasks.

Large repositories of image-caption pairs are essential for the development of vision-language models. However, these datasets are often extracted from noisy data scraped from the web, and contain many mislabeled instances. In order to improve the reliability of downstream models, it is important to identify and filter images with incorrect captions. However, beyond filtering based on image-caption embedding similarity, no prior works have proposed other methods to filter noisy multimodal data, or concretely assessed the impact of noisy captioning data on downstream training. In this work, we propose, theoretically justify, and empirically validate LEMoN, a method to identify label errors in image-caption datasets. Our method leverages the multimodal neighborhood of image-caption pairs in the latent space of contrastively pretrained multimodal models to automatically identify label errors. Through empirical evaluations across eight datasets and twelve baselines, we find that LEMoN outperforms the baselines by over 3% in label error detection, and that training on datasets filtered using our method improves downstream captioning performance by more than 2 BLEU points over noisy training.

Health AI suffers from a systemic reproducibility crisis that irreparably hinders research across both academia and industry [4,5].One key tool poised to solve this crisis is the Medical Event Data Standard (MEDS), a comprehensive data format and open-source ecosystem designed to enhance reproducibility and interoperability of AI research using longitudinal Electronic Health Records (EHR) [6].Currently adopted by over 15 institutions globally, MEDS encompasses various open-source tools, published models, and data processing pipelines, enabling streamlined model development and robust benchmarking.In this tutorial, participants will gain key hands-on experience in working with the MEDS format to perform efficient, reproducible, state-of-the-art AI research over real health data.Participants will transform data into the MEDS format, preprocess data, build predictive models, and contribute to the decentralized MEDS-DEV benchmarking platform.Interactive exercises using Jupyter notebooks will provide hands-on experience and practical skills for reproducible health AI research.Attendees will leave equipped

Foundation models are becoming valuable tools in medicine. Yet despite their promise, the best way to leverage Large Language Models (LLMs) in complex medical tasks remains an open question. We introduce a novel multi-agent framework, named **M**edical **D**ecision-making **Agents** (**MDAgents**) that helps to address this gap by automatically assigning a collaboration structure to a team of LLMs. The assigned solo or group collaboration structure is tailored to the medical task at hand, a simple emulation inspired by the way real-world medical decision-making processes are adapted to tasks of different complexities. We evaluate our framework and baseline methods using state-of-the-art LLMs across a suite of real-world medical knowledge and clinical diagnosis benchmarks, including a comparison of LLMs’ medical complexity classification against human physicians. MDAgents achieved the **best performance in seven out of ten** benchmarks on tasks requiring an understanding of medical knowledge and multi-modal reasoning, showing a significant **improvement of up to 4.2\%** ($p$ < 0.05) compared to previous methods' best performances. Ablation studies reveal that MDAgents effectively determines medical complexity to optimize for efficiency and accuracy across diverse medical tasks. Notably, the combination of moderator review and external medical knowledge in group collaboration resulted in an average accuracy **improvement of 11.8\%**. Our code can be found at https://github.com/mitmedialab/MDAgents.

The proliferation of deep learning-based machine vision applications has given rise to a new type of compression, so called video coding for machine (VCM). VCM differs from traditional video coding in that it is optimized for machine vision performance instead of human visual quality. In the feature compression track of MPEG-VCM, multi-scale features extracted from images are subject to compression. Recent feature compression works have demonstrated that the versatile video coding (VVC) standard-based approach can achieve a BD-rate reduction of up to 96% against MPEG-VCM feature anchor. However, it is still sub-optimal as VVC was not designed for extracted features but for natural images. Moreover, the high encoding complexity of VVC makes it difficult to design a lightweight encoder without sacrificing performance. To address these challenges, we propose a novel multi-scale feature compression method that enables both the end-to-end optimization on the extracted features and the design of lightweight encoders. The proposed model combines a learnable compressor with a multi-scale feature fusion network so that the redundancy in the multi-scale features is effectively removed. Instead of simply cascading the fusion network and the compression network, we integrate the fusion and encoding processes in an interleaved way. Our model first encodes a larger-scale feature to obtain a latent representation and then fuses the latent with a smaller-scale feature. This process is successively performed until the smallest-scale feature is fused and then the encoded latent at the final stage is entropy-coded for transmission. The results show that our model outperforms previous approaches by at least 52% BD-rate reduction and has$\times 5$to$\times 27$times less encoding time for object detection. It is noteworthy that our model can attain near-lossless task performance with only 0.002-0.003% of the uncompressed feature data size.

In search for efficient feature compression technologies for machine consumption, MPEG recently issued a call for proposal (CfP) on feature compression for video coding for machine (FCVCM). One issue in feature compression is that the input feature maps generally have high redundancy in them. Various researches to reduce such redundancy have been made. For example, a recent study called L-MSFC (learnable multi-scale feature compression), which effectively combines multi-scale feature fusion and compression in an end-to-end learnable framework, showed up to 98% BD rate gain over the anchor model defined in the FCVCM CfP. Despite these advances in FCVCM, relation between distortions in feature maps and performance of vision tasks has stayed relatively unexplored. In this paper, we propose a novel loss function called MEDO (minimizing effective distortions only) based on our hypothesis that distortions below some threshold do not improve task performance. Experimental results on instance segmentation task show that our MEDO loss on top of L-MSFC improves the overall rate-mAP performance without compromising complexity. Being more practical for real-world uses, we also present an extension to L-MSFC for variable-rate support with a single model.

Although recent multi-task learning methods have shown to be effective in improving the generalization of deep neural networks, they should be used with caution for safety-critical applications, such as clinical risk prediction. This is because even if they achieve improved task-average performance, they may still yield degraded performance on individual tasks, which may be critical (e.g., prediction of mortality risk). Existing asymmetric multi-task learning methods tackle this negative transfer problem by performing knowledge transfer from tasks with low loss to tasks with high loss. However, using loss as a measure of reliability is risky since low loss could result from overfitting. In the case of time-series prediction tasks, knowledge learned for one task (e.g., predicting the sepsis onset) at a specific timestep may be useful for learning another task (e.g., prediction of mortality) at a later timestep, but lack of loss at each timestep makes it difficult to measure the reliability at each timestep. To capture such dynamically changing asymmetric relationships between tasks in time-series data, we propose a novel temporal asymmetric multi-task learning model that performs knowledge transfer from certain tasks/timesteps to relevant uncertain tasks, based on the feature-level uncertainty. We validate our model on multiple clinical risk prediction tasks against various deep learning models for time-series prediction, which our model significantly outperforms without any sign of negative transfer. Further qualitative analysis of learned knowledge graphs by clinicians shows that they are helpful in analyzing the predictions of the model.

We propose a novel interactive learning framework which we refer to as Interactive Attention Learning (IAL), in which the human supervisors interactively manipulate the allocated attentions, to correct the model’s behaviour by updating the attention-generating network. However, such a model is prone to overfitting due to scarcity of human annotations, and requires costly retraining. Moreover, it is almost infeasible for the human annotators to examine attentions on tons of instances and features. We tackle these challenges by proposing a sample-efficient attention mechanism and a cost-effective reranking algorithm for instances and features. First, we propose Neural Attention Processes (NAP), which is an attention generator that can update its behaviour by incorporating new attention-level supervisions without any retraining. Secondly, we propose an algorithm which prioritizes the instances and the features by their negative impacts, such that the model can yield large improvements with minimal human feedback. We validate IAL on various time-series datasets from multiple domains (healthcare, real-estate, and computer vision) on which it significantly outperforms baselines with conventional attention mechanisms, or without cost-effective reranking, with substantially less retraining and human-model interaction cost.

A new architecture for matching the data protected with an error-correcting code (ECC) is presented in this brief to reduce latency and complexity. Based on the fact that the codeword of an ECC is usually represented in a systematic form consisting of the raw data and the parity information generated by encoding, the proposed architecture parallelizes the comparison of the data and that of the parity information. To further reduce the latency and complexity, in addition, a new butterfly-formed weight accumulator (BWA) is proposed for the efficient computation of the Hamming distance. Grounded on the BWA, the proposed architecture examines whether the incoming data matches the stored data if a certain number of erroneous bits are corrected. For a (40, 33) code, the proposed architecture reduces the latency and the hardware complexity by ~32% and 9%, respectively, compared with the most recent implementation.