EffiMiniVLM is a 27.7M-parameter vision-language regression model that reaches a CES score of 0.40 on Amazon Reviews 2023 data while using 6.8 GFLOPs and no external datasets.
Tri-axial scaling in aerial object detec- tion: Model size, dataset size and quality, and test-time in- ference in the cadot challenge
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
EffiMiniVLM: A Compact Dual-Encoder Regression Framework
EffiMiniVLM is a 27.7M-parameter vision-language regression model that reaches a CES score of 0.40 on Amazon Reviews 2023 data while using 6.8 GFLOPs and no external datasets.