Finetuning on the Crowd Human dataset to improve people detection

[simpad2409]

Hello everyone and thanks in advance to those who decide to help me. I will be really grateful to him. I am working on my master's thesis and I would like to perform a finetuning on the Crowd Human dataset, in order to improve the detection of people.
But I think I'm doing something wrong ... The model I get is bad! The results on the test set are much worse than the original DETR network.

Could anyone help me?

Thank you so much. I would be immensely grateful.

(I could attach the code, although I followed that guide step by step:
Link
modifying only the dataset (train and validation) and the two annotation files.

Please help me. Thanks in advance.

[ni3s]

comment here if you still need any help