Behind the Magic, MERLIM: Multi-modal Evaluation Benchmark for Large Image-Language Models.
Andrés VillaJuan Carlos León AlcázarAlvaro SotoBernard GhanemPublished in: CoRR (2023)
Keyphrases
- multi modal
- language model
- uni modal
- image data
- image features
- language modeling
- speech recognition
- image analysis
- probabilistic model
- n gram
- image content
- image classification
- document retrieval
- image segmentation
- multi modality
- single modality
- retrieval model
- audio visual
- image retrieval
- multiple modalities
- statistical language models
- image representation
- language modelling
- information retrieval
- image collections
- cross modal
- high dimensional
- image regions
- low level
- video search
- document ranking