"float division by zero" when I validate my own dataset

[zxq-creations]

I'm trying to validate my own dataset with pairs.txt, but I met such a headache issue. Here is the src of my issue: I got my own dataset and pairs.txt, aligned them and now I run "validate_on_lfw.py". But after a long waiting, I get a Traceback

Traceback (most recent call last): File "facenet/src/validate_on_pcs.py", line 117, in <module> main(parse_arguments(sys.argv[1:])) File "facenet/src/validate_on_pcs.py", line 85, in main actual_issame, nrof_folds=args.lfw_nrof_folds) File "/root/prj_facenet/facenet/src/pcs.py", line 43, in evaluate np.asarray(actual_issame), 1e-3, nrof_folds=nrof_folds) File "/root/prj_facenet/facenet/src/facenet.py", line 489, in calculate_val val[fold_idx], far[fold_idx] = calculate_val_far(threshold, dist[test_set], actual_issame[test_set]) File "/root/prj_facenet/facenet/src/facenet.py", line 504, in calculate_val_far far = float(false_accept) / float(n_diff) ZeroDivisionError: float division by zero

By review the src, I found the zero is created here(in facenet.py: 501L): def calculate_val_far(threshold, dist, actual_issame): predict_issame = np.less(dist, threshold) true_accept = np.sum(np.logical_and(predict_issame, actual_issame)) false_accept = np.sum(np.logical_and(predict_issame, np.logical_not(actual_issame))) n_same = np.sum(actual_issame) n_diff = np.sum(np.logical_not(actual_issame)) val = float(true_accept) / float(n_same) far = float(false_accept) / float(n_diff) return val, far

the n_diff is valued 0 as the actual_diff here is 'True'. So it seems like there is no logical error, but when validate lfw dataset, it runs very fluently with no barrier. PUZZLING......CRYING....... Thx for any help from any buddy.

[LiuNull]

change your pair.txt, maybe it is because there is no different people sample in your batch

[abrass]

same problem here with that pairs.txt Abel_Pacheco 1 4 Akhmed_Zakayev 1 3 Akhmed_Zakayev 2 3 Amber_Tamblyn 1 2 Robby_Ginepri 1 2 Robert_Witt 1 2 Rod_Blagojevich 1 2 Rolandas_Paksas 1 2 Russell_Coutts 1 2 Ruth_Dreifuss 1 2 Sally_Kirkland 1 4

[nadaboulares]

@abrass how can i correct my file