Year: 2026
Type: preprint
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 224x224 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: arXiv (Cornell University)
Institutions: Columbia University, University of Pennsylvania
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funders: Foundation for the National Institutes of Health, Alzheimer's Association, Bristol-Myers Squibb, Eli Lilly and Company, Pfizer, Massachusetts General Hospital, Biogen, University of Southern California, F. Hoffmann-La Roche, Meso Scale Diagnostics, Harvard University, Alzheimer's Disease Neuroimaging Initiative, BioClinica, Novartis Pharmaceuticals Corporation, Northern California Institute for Research and Education, Eisai, Southwest University, Centre d'Imagerie BioMédicale, Servier, H. Lundbeck A/S, IXICO, National Institutes of Health, Genentech, Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Canadian Institutes of Health Research, National Institute on Aging, National Institute of Mental Health, National Center for Research Resources
Awards: G1156091317, Mapping Abnormal Neurodevelopment in Schizophrenia, Functional-anatomic exploration of cognitive control, G2502532649, Morphometry Biomedical Informatics Research Network, G4079001475, G5047191974, G5106486648, G5235017436, G6452318992, Alzheimer's Disease Research Center, Healthy Aging and Senile Dementia, SchizConnect: Large-Scale Schizophrenia Neuroimaging Data Mediation & Federation, G8448649937, G8986923329, G9000197540, Alzheimer's Disease Neuroimaging Initiative (ADNI4), Antecedent Biomarkers for Alzheimer Disease: The Adult Children Study
OpenAlex ID: W7220940296