Prepares a grid to use with grid_sample to upsample a batch of features in data_shape to the target_shape. Can use scale_offset and input_shap

Sets up a set of layers to be hooked. Usually not called directly: use edit_layer or retain_layer instead.

Input: set of rows with 'number' fields (or another field name key). Output: array such that a[number] = the row with the given number.

Returns maxiou and maxiou_level across the data set, one per layer. This is a performance-sensitive function. Best performance is achie

Splits connected components into regions (slower, requires cpu).

Runs net dissection in-memory, using pytorch, and saves visualizations and metadata into outdir.