iFormer combines a mobile-tuned ConvNeXt backbone with single-head modulation attention, reaching 80.4% ImageNet top-1 accuracy at 1.10 ms iPhone 13 latency.
As summarized in Table 14, split and concatenate operations introduce additional runtime
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
iFormer: Integrating ConvNet and Transformer for Mobile Application
iFormer combines a mobile-tuned ConvNeXt backbone with single-head modulation attention, reaching 80.4% ImageNet top-1 accuracy at 1.10 ms iPhone 13 latency.