AI Insight
Researchers developed FlyVision, a computer vision architecture inspired by the Drosophila (fruit fly) visual connectome, which achieves competitive image classification performance while using significantly fewer parameters than conventional models. The architecture preserved key biological features including parallel ON/OFF processing and recurrent computation, reaching accuracies within three percentage points of ResNet18 on ImageNet-1K while using only 16-32% of the parameters. The model successfully scaled from basic digit recognition to complex medical imaging tasks including skin disease classification, chest radiography, and brain age estimation from MRI scans.
Why it matters
This work demonstrates that biological neural architectures can provide efficient computational principles for artificial intelligence, potentially reducing the computational resources required for deploying computer vision models in real-world applications. The successful application to medical imaging tasks, including brain age prediction with high accuracy, suggests practical utility for resource-constrained clinical settings.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Abstract: Biological connectomes encode structured solutions to visual computation that may provide reusable inductive biases for artificial vision. We develop ConnectomeX around FlyVision, a trainable architecture that preserves parallel ON/OFF processing, recurrent computation and population-level graph interaction while scaling model capacity across tasks. FlyVision reached 99.34% accuracy on MNIST with 80,608 parameters and 78.03% on CIFAR-10 with 81,408 parameters. On ImageNet-1K, FlyVision Base and Large reached 60.79% and 66.25% top-1 accuracy with 1.8 and 3.7 million parameters, while a Large local-k7 model with a learned low-frequency branch reached 66.53%, compared with 69.25% for ResNet18 with 11.7 million parameters. On a 22-class skin-disease benchmark, FlyVision Large achieved 63.78% accuracy and 95.28% macro-AUROC with 2.99 million parameters. In four-class chest radiography, ImageNet-pretrained FlyVision Base and Large reached 92.60% and 92.76% accuracy with 1.33 and 2.97 million parameters, compared with 91.56% for ImageNet-pretrained ResNet18 with 11.18 million. BrainAGE extends FlyVision to volumetric T1-weighted MRI by applying a shared ImageNet-pretrained FlyVision Large encoder to 24 sagittal, coronal and axial slices per scan and combining slice-level age estimates by confidence-modulated Gaussian voting. On 433 held-out scans, three-axis fusion achieved a mean absolute error of 5.98 years and R^2 = 0.868. Across the 224×224 classification tasks, the best FlyVision configuration remained within three percentage points of ResNet18 on ImageNet-1K and skin-disease classification and exceeded it on chest radiography with substantially fewer parameters. These results show that a conserved connectome-informed computation can scale from compact recognition to large-scale natural and biomedical vision.
Source: From the Drosophila Visual Connectome to General-Purpose Computer Vision