arXiv ·
Researchers at MBZUAI have introduced MedMerge, a transfer learning technique that merges weights from independently initialized models to improve performance on medical imaging tasks. MedMerge learns kernel-level weights to combine features from different models into a single model. Experiments across various medical imaging tasks demonstrated performance gains of up to 7% in F1 score.
arXiv ·
Researchers at MBZUAI introduce FissionFusion, a hierarchical model merging approach to improve medical image analysis performance. The method uses local and global aggregation of models based on hyperparameter configurations, along with a cyclical learning rate scheduler for efficient model generation. Experiments show FissionFusion outperforms standard model souping by approximately 6% on HAM10000 and CheXpert datasets and improves OOD performance.
arXiv ·
Researchers at MBZUAI have developed DynaMMo, a dynamic model merging method for efficient class incremental learning using medical images. DynaMMo merges multiple networks at different training stages using lightweight learnable modules, reducing computational overhead. Evaluated on three datasets, DynaMMo achieved a 10-fold reduction in GFLOPS compared to existing dynamic methods with a 2.76 average accuracy drop.
arXiv ·
MBZUAI researchers have introduced MIRA, a novel framework for improving the factual accuracy of multimodal large language models in medical applications. MIRA uses calibrated retrieval to manage factual risk and integrates image embeddings with a medical knowledge base for efficient reasoning. Evaluated on medical VQA and report generation benchmarks, MIRA achieves state-of-the-art results, with code available on GitHub.
arXiv ·
The paper introduces MedNNS, a neural network search framework designed for medical imaging, addressing challenges in architecture selection and weight initialization. MedNNS constructs a meta-space encoding datasets and models based on their performance using a Supernetwork-based approach, expanding the model zoo size by 51x. The framework incorporates rank loss and Fréchet Inception Distance (FID) loss to capture inter-model and inter-dataset relationships, improving alignment in the meta-space and outperforming ImageNet pre-trained DL models and SOTA NAS methods.