MemeBLIP2, built on BLIP-2 with linear projections, adapters, and an MLP classifier, reaches 77.5% accuracy and 79.0% F1 on PrideMM harmful meme detection, but the paper's internal inconsistencies weaken the claim.
HateCLIPper: A CLIP- based Framework for Multimodal Hate Speech Detec- tion
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
-
MemeBLIP2: A novel lightweight multimodal system to detect harmful memes
MemeBLIP2, built on BLIP-2 with linear projections, adapters, and an MLP classifier, reaches 77.5% accuracy and 79.0% F1 on PrideMM harmful meme detection, but the paper's internal inconsistencies weaken the claim.