search

DAMO-NLP-SG / VCD

[CVPR 2024 Highlight] Mitigating Object Hallucinations in Large Vision-Language Models through Visual Contrastive Decoding
Apache License 2.0
216 stars 10 forks source link

Ask for parameters #26

Open maming109 opened 1 day ago

maming109 commented 1 day ago

Hello author, can you tell me about the relevant parameters of your experiment, the results I have made are very poor, best wishes！ Table 3. Results of GPT-4V-aided evaluation on open-ended generation. Accuracy measures the response’s alignment with the image content, and Detailedness gauges the richness of details in the response. Both metrics are on a scale of 10.

LengSicong commented 1 day ago

The hyperparameters are provided in the main paper and appendix.