Adapting Grounded Visual Question Answering Models to Low Resource Languages

Computer Science

• Contrastive Loss 50%
• Embedding Space 50%
• Language Modeling 100%
• Language Resource 100%
• Popular Approach 50%
• translation quality 50%
• Visual Question Answering 100%

Keyphrases

• Modular Training 50%
• Multilingual Representations 50%
• Multimodal Model 50%
• Pre-trained Weights 50%
• Vision-language Models 50%

Social Sciences

• Bengali 50%
• Language Modeling 100%
• Modular Instruction 50%