You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello, thank you for the interesting work. Though I have a question about the vqa score calculation. When I run the eval.py to calculate the evaluation scores for vqa, I find the min(normlized_answers.get(vqa_preprocess(answer), 0) / 3, 1) is used to calculate the score (accuracy?). However, I am confused by the /3 here. Can you explain this to me thanks! def vqa_score(answer, ground_truth_answer_counts) in evaluator.py
The text was updated successfully, but these errors were encountered:
Hello, thank you for the interesting work. Though I have a question about the vqa score calculation. When I run the eval.py to calculate the evaluation scores for vqa, I find the min(normlized_answers.get(vqa_preprocess(answer), 0) / 3, 1) is used to calculate the score (accuracy?). However, I am confused by the /3 here. Can you explain this to me thanks!
def vqa_score(answer, ground_truth_answer_counts)
in evaluator.pyThe text was updated successfully, but these errors were encountered: