Fix test fail: Due to recompute_scale_factor missing in UpsamplingBilinear2d

[kingjuno]

There are issues in using UpsamplingBilinear2d for torch version >= 1.9.0 [ref]

One possible solution is to edit ..site-packages\torch\nn\modules:

def forward(self, input: Tensor) -> Tensor:
    return F.interpolate(input, self.size, self.scale_factor, self.mode, self.align_corners,
                         #recompute_scale_factor=self.recompute_scale_factor
                         )

But this isn't possible in this case. So this PR is to manually set recompute_scale_factor.

[codecov-commenter]

Codecov Report

Merging #68 (c23e4da) into master (d5baccc) will increase coverage by 0.06%. The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #68      +/-   ##
==========================================
+ Coverage   86.15%   86.22%   +0.06%     
==========================================
  Files          12       12              
  Lines         419      421       +2     
==========================================
+ Hits          361      363       +2     
  Misses         58       58              

Impacted Files	Coverage Δ
...cell_membrane_segmentor/cell_membrane_segmentor.py	87.17% <100.00%> (+0.11%)	:arrow_up:
...n/cell_nucleus_segmentor/cell_nucleus_segmentor.py	89.28% <100.00%> (+0.19%)	:arrow_up:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 904f454...c23e4da. Read the comment docs.

[Mayukhdeb]

Thanks @kingjuno for the PR ^^ I've left a comment in one of the diffs, please take a look.

[Mayukhdeb]

Thanks @kingjuno again for the PR!