Key-Frame Detection and Super-Resolution of Hyperspectral Video via Sparse-Based Cumulative Tensor Factorization

Table 1

Detection quantitative results (AUC value) of the test methods on key-frames.

Frame	MSD [25]	CMF [29]	CTCF
12	0.9655	0.9993	0.9980
13	0.8462	0.9995	0.9980
14	0.8878	0.9994	0.9981
15	0.8189	0.9965	0.9981
16	0.8734	0.9995	0.9977
17	0.9348	0.9987	0.9946
18	0.5792	0.7477	0.9915
19	0.7894	0.9988	0.9958
20	0.9336	0.8991	0.9934
21	0.8222	0.9980	0.9966
22	0.9001	0.9915	0.9969
23	0.8388	0.9986	0.9945
24	0.8914	0.9990	0.9983
25	0.9169	0.9989	0.9961
26	0.9254	0.9947	0.9989
27	0.8722	0.9974	0.9951
28	0.8503	0.9978	0.9930
29	0.9490	0.9892	0.9974
30	0.9011	0.9341	0.9885
31	0.9157	0.9867	0.9965
32	0.8881	0.8582	0.9811
33	0.9345	0.9771	0.9933
34	0.9007	0.9977	0.9922
35	0.9273	0.9952	0.9933
36	0.9349	0.9989	0.9950
37	0.9528	0.9986	0.9984
38	0.9299	0.9875	0.9981
39	0.8838	0.9979	0.9962
40	0.9295	0.9976	0.9966
41	0.9165	0.9939	0.9976
42	0.9665	0.9986	0.9988
43	0.9660	0.9763	0.9969
44	0.9083	0.9995	0.9979
45	0.9046	0.9977	0.9935
46	0.9156	0.9964	0.9918
47	0.9225	0.9979	0.9950
48	0.8623	0.9980	0.9967
49	0.8894	0.9992	0.9980
50	0.8640	0.9969	0.9947
Average	0.8926	0.9817	0.9954
Variance	0.4407 × 10^−2	0.2290 × 10^−2	0.1134 × 10^−6