The Saliency Sampler is not used in the experiment

[kevinlemon]

Thanks for your code. (1) In the parse_args.py, the saliency sampler module is defaulted as 0. There are no additional commands to change the default value in the experiment (both "Training and Testing" and "Running Batch Experiments").

Saliency sampler module

    ss_parser = parser.add_argument_group('Saliency Sampler')
    ss_parser.add_argument('--ss', type=int, default=0, help='toggle saliency sampler [0,1]')
    ss_parser.add_argument('--ss_pretrain', type=int, default=0, help='toggle saliency sampler ImageNet pretraining')
    ss_parser.add_argument('--ss_dim', default=None, type=int, nargs='+', help='width and height of saliency network input in pixels')
    ss_parser.add_argument('--ss_out_dim', default=None, type=int, nargs='+', help='width and height of saliency network input in pixels')
    ss_parser.add_argument('--ss_layers', default=None, type=int, help='number of layers to include in saliency net')
    ss_parser.add_argument('--ss_sparsity', default=0, type=float, help='weighting for sparsity loss')
    ss_parser.add_argument('--ss_temporal', default=0, type=float, help='weighting for temporal consistency loss')

(2) There are ppg, our, pred_ in the results. What are the meanings of these prefix? test_Last ppg_NegMCC: -0.5935 test_Last ppg_NegSNR: -14.5025 test_Last ppg_IPR: 0.4602 test_Last ppg_NegPC: 0.3113 test_Last ppg_MVTL: 0.0000 test_Last our_v_gt_sd: 5.2483 test_Last our_v_gt_rmse: 6.1034 test_Last our_v_gt_mae: 4.6451 test_Last our_v_gt_corr: 0.8769 test_Last pred_v_gt_sd: 6.2129 test_Last pred_v_gt_rmse: 6.9774 test_Last pred_v_gt_mae: 5.2871 test_Last pred_v_gt_corr: 0.8370 test_Last pred_v_our_sd: 1.1367 test_Last pred_v_our_rmse: 1.1382 test_Last pred_v_our_mae: 0.8260 test_Last pred_v_our_corr: 0.9955

[lbh111w]

I also have this question (2).

[Mollyzml]

I also have this question (2).