0c63b9dcd7ee3da39c74379a01d16e3a360d3506,niftynet/layer/resampler.py,ResamplerLayer,_resample_inv_dst_weighting,#ResamplerLayer#Any#Any#,178

Before Change

            self.boundary_func(tf.ceil(sample_coords), b_size)], axis=0)

        // find N weights associated to each output point
        all_coords_f = tf.to_float(all_coords)
        diff = tf.stack(
            [tf.squared_difference(sample_coords, all_coords_f[0]),
             tf.squared_difference(sample_coords, all_coords_f[1])])
        point_weights = tf.gather_nd(diff, weight_id)
        point_weights = tf.reduce_sum(point_weights, axis=1)
        // skip this as power = 2:
        // self.power = 2
        // point_weights = tf.pow(point_weights, self.power / 2.0)

After Change

        // gather_nd for both matrix, the same as:
        // point_weights = tf.gather_nd(diff, weight_id)
        // knots_id = tf.gather_nd(all_coords_f, weight_id)
        n_val = tf.gather_nd(tf.stack([diff, all_coords_f], axis=-1), weight_id)
        n_val = tf.unstack(n_val, axis=-1)
        point_weights, knots_id = n_val[0], n_val[1]

        point_weights = tf.reduce_sum(point_weights, axis=1)

Instances