FastVim
A lightweight image classification model based on the vision Mamba architecture that can distill pre-trained weights from large foundation models, substantially reducing parameter count and inference cost while retaining semantic knowledge.
Timeline (last 90 days)
The method uses one Mamba model at each scale to encode features from two sources—a frozen foundation model and the lightweight FastVim—in linear time, after which a jointly trained classification head fuses the encoded representations into a single prediction.
FastVim is a lightweight variant based on the vision Mamba architecture, designed specifically for efficient image classification, and its pretrained weights can be obtained by distillation from a large foundation model.
All Facts (2)
The method uses one Mamba model at each scale to encode features from two sources—a frozen foundation model and the lightweight FastVim—in linear time, after which a jointly trained classification head fuses the encoded representations into a single prediction.
50%UnverifiedFastVim is a lightweight variant based on the vision Mamba architecture, designed specifically for efficient image classification, and its pretrained weights can be obtained by distillation from a large foundation model.
50%