DCT stats list -> tuple

Author: rwightman

timm/models/csatv2.py

@@ -3,6 +3,12 @@
 A frequency-domain vision model using DCT transforms with spatial attention.
 
 Paper: TBD
+
+This model created by members of MLPA Lab. Welcome feedback and suggestion, questions.
+[email protected]
+[email protected]
+
+Refined for timm by Ross Wightman
 """
 import math
 import warnings
@@ -24,59 +30,59 @@
 __all__ = ['CSATv2', 'csatv2']
 
 # DCT frequency normalization statistics (Y, Cb, Cr channels x 64 coefficients)
-_DCT_MEAN = [
-    [932.42657, -0.00260, 0.33415, -0.02840, 0.00003, -0.02792, -0.00183, 0.00006,
+_DCT_MEAN = (
+    (932.42657, -0.00260, 0.33415, -0.02840, 0.00003, -0.02792, -0.00183, 0.00006,
      0.00032, 0.03402, -0.00571, 0.00020, 0.00006, -0.00038, -0.00558, -0.00116,
      -0.00000, -0.00047, -0.00008, -0.00030, 0.00942, 0.00161, -0.00009, -0.00006,
      -0.00014, -0.00035, 0.00001, -0.00220, 0.00033, -0.00002, -0.00003, -0.00020,
      0.00007, -0.00000, 0.00005, 0.00293, -0.00004, 0.00006, 0.00019, 0.00004,
      0.00006, -0.00015, -0.00002, 0.00007, 0.00010, -0.00004, 0.00008, 0.00000,
      0.00008, -0.00001, 0.00015, 0.00002, 0.00007, 0.00003, 0.00004, -0.00001,
-     0.00004, -0.00000, 0.00002, -0.00000, -0.00008, -0.00000, -0.00003, 0.00003],
-    [962.34735, -0.00428, 0.09835, 0.00152, -0.00009, 0.00312, -0.00141, -0.00001,
+     0.00004, -0.00000, 0.00002, -0.00000, -0.00008, -0.00000, -0.00003, 0.00003),
+    (962.34735, -0.00428, 0.09835, 0.00152, -0.00009, 0.00312, -0.00141, -0.00001,
      -0.00013, 0.01050, 0.00065, 0.00006, -0.00000, 0.00003, 0.00264, 0.00000,
      0.00001, 0.00007, -0.00006, 0.00003, 0.00341, 0.00163, 0.00004, 0.00003,
      -0.00001, 0.00008, -0.00000, 0.00090, 0.00018, -0.00006, -0.00001, 0.00007,
      -0.00003, -0.00001, 0.00006, 0.00084, -0.00000, -0.00001, 0.00000, 0.00004,
      -0.00001, -0.00002, 0.00000, 0.00001, 0.00002, 0.00001, 0.00004, 0.00011,
      0.00000, -0.00003, 0.00011, -0.00002, 0.00001, 0.00001, 0.00001, 0.00001,
-     -0.00007, -0.00003, 0.00001, 0.00000, 0.00001, 0.00002, 0.00001, 0.00000],
-    [1053.16101, -0.00213, -0.09207, 0.00186, 0.00013, 0.00034, -0.00119, 0.00002,
+     -0.00007, -0.00003, 0.00001, 0.00000, 0.00001, 0.00002, 0.00001, 0.00000),
+    (1053.16101, -0.00213, -0.09207, 0.00186, 0.00013, 0.00034, -0.00119, 0.00002,
      0.00011, -0.00984, 0.00046, -0.00007, -0.00001, -0.00005, 0.00180, 0.00042,
      0.00002, -0.00010, 0.00004, 0.00003, -0.00301, 0.00125, -0.00002, -0.00003,
      -0.00001, -0.00001, -0.00001, 0.00056, 0.00021, 0.00001, -0.00001, 0.00002,
      -0.00001, -0.00001, 0.00005, -0.00070, -0.00002, -0.00002, 0.00005, -0.00004,
      -0.00000, 0.00002, -0.00002, 0.00001, 0.00000, -0.00003, 0.00004, 0.00007,
      0.00001, 0.00000, 0.00013, -0.00000, 0.00000, 0.00002, -0.00000, -0.00001,
-     -0.00004, -0.00003, 0.00000, 0.00001, -0.00001, 0.00001, -0.00000, 0.00000],
-]
+     -0.00004, -0.00003, 0.00000, 0.00001, -0.00001, 0.00001, -0.00000, 0.00000),
+)
 
-_DCT_VAR = [
-    [270372.37500, 6287.10645, 5974.94043, 1653.10889, 1463.91748, 1832.58997, 755.92468, 692.41528,
+_DCT_VAR = (
+    (270372.37500, 6287.10645, 5974.94043, 1653.10889, 1463.91748, 1832.58997, 755.92468, 692.41528,
      648.57184, 641.46881, 285.79288, 301.62100, 380.43405, 349.84027, 374.15891, 190.30960,
      190.76746, 221.64578, 200.82646, 145.87979, 126.92046, 62.14622, 67.75562, 102.42001,
      129.74922, 130.04631, 103.12189, 97.76417, 53.17402, 54.81048, 73.48712, 81.04342,
      69.35100, 49.06024, 33.96053, 37.03279, 20.48858, 24.94830, 33.90822, 44.54912,
      47.56363, 40.03160, 30.43313, 22.63899, 26.53739, 26.57114, 21.84404, 17.41557,
      15.18253, 10.69678, 11.24111, 12.97229, 15.08971, 15.31646, 8.90409, 7.44213,
-     6.66096, 6.97719, 4.17834, 3.83882, 4.51073, 2.36646, 2.41363, 1.48266],
-    [18839.21094, 321.70932, 300.15259, 77.47830, 76.02293, 89.04748, 33.99642, 34.74807,
+     6.66096, 6.97719, 4.17834, 3.83882, 4.51073, 2.36646, 2.41363, 1.48266),
+    (18839.21094, 321.70932, 300.15259, 77.47830, 76.02293, 89.04748, 33.99642, 34.74807,
      32.12333, 28.19588, 12.04675, 14.26871, 18.45779, 16.59588, 15.67892, 7.37718,
      8.56312, 10.28946, 9.41013, 6.69090, 5.16453, 2.55186, 3.03073, 4.66765,
      5.85418, 5.74644, 4.33702, 3.66948, 1.95107, 2.26034, 3.06380, 3.50705,
      3.06359, 2.19284, 1.54454, 1.57860, 0.97078, 1.13941, 1.48653, 1.89996,
      1.95544, 1.64950, 1.24754, 0.93677, 1.09267, 1.09516, 0.94163, 0.78966,
      0.72489, 0.50841, 0.50909, 0.55664, 0.63111, 0.64125, 0.38847, 0.33378,
-     0.30918, 0.33463, 0.20875, 0.19298, 0.21903, 0.13380, 0.13444, 0.09554],
-    [17127.39844, 292.81421, 271.45209, 66.64056, 63.60253, 76.35437, 28.06587, 27.84831,
+     0.30918, 0.33463, 0.20875, 0.19298, 0.21903, 0.13380, 0.13444, 0.09554),
+    (17127.39844, 292.81421, 271.45209, 66.64056, 63.60253, 76.35437, 28.06587, 27.84831,
      25.96656, 23.60370, 9.99173, 11.34992, 14.46955, 12.92553, 12.69353, 5.91537,
      6.60187, 7.90891, 7.32825, 5.32785, 4.29660, 2.13459, 2.44135, 3.66021,
      4.50335, 4.38959, 3.34888, 2.97181, 1.60633, 1.77010, 2.35118, 2.69018,
      2.38189, 1.74596, 1.26014, 1.31684, 0.79327, 0.92046, 1.17670, 1.47609,
      1.50914, 1.28725, 0.99898, 0.74832, 0.85736, 0.85800, 0.74663, 0.63508,
      0.58748, 0.41098, 0.41121, 0.44663, 0.50277, 0.51519, 0.31729, 0.27336,
-     0.25399, 0.27241, 0.17353, 0.16255, 0.18440, 0.11602, 0.11511, 0.08450],
-]
+     0.25399, 0.27241, 0.17353, 0.16255, 0.18440, 0.11602, 0.11511, 0.08450),
+)
 
 
 def _zigzag_permutation(rows: int, cols: int) -> List[int]: