in demo.py, amodel_ct_output should not be affected by reg value?

[likegogogo]

according to the src\lib\dataset\generic_dataset.py code: hm ->ct_int ret['reg'][k] = ct - ct_int ret['amodel_offset'][k] = amodel_center - ct_int

so, in inference, should be: ct_int: hm ct: ret['reg'][k] + ct_int = ret['reg'][k] + hm amodel_center: ret['amodel_offset'][k] + ct_int = ret['amodel_offset'][k] + hm

in training, reg value is not used when calc amodel_offset. But in inference, when calc bboxes, the code used reg values. So there is a gap??

    if 'amodel_offset' in dets and len(dets['amodel_offset'][i]) > j:
      ct_output = dets['bboxes'][i][j].reshape(2, 2).mean(axis=0)
      amodel_ct_output = ct_output + dets['amodel_offset'][i][j]
      ct = transform_preds_with_trans(
        amodel_ct_output.reshape(1, 2), trans).reshape(2).tolist()