Abstract: Systems and techniques are provided for processing image data. For example, a process can include receiving an input image comprising a padded image including a plurality of padding pixels located around a perimeter of the frame of image data. A plurality of overlapping image patches can be generated, each corresponding to a respective portion of the padded image and having a resolution. A channel-wise grouped input can be generated from the plurality of overlapping image patches and having the resolution, and includes a corresponding channel for each respective channel of each overlapping image patch of the plurality of overlapping image patches. An image processing output can be generated based on processing the channel-wise grouped input using a plurality of grouped convolutional layers of a tiled neural network, wherein at least one grouped convolutional layer of the plurality of grouped convolutional layers does not perform internal padding.

Abstract: Ladle for the filtration of a liquid metal such as aluminum, magnesium, and their alloys over a filter medium of the gravel bed type. The ladle comprises a chamber in which the filter medium rests on a supporting grid, an inlet for supplying liquid metal to the chamber and an outlet for removing liquid metal from the chamber. The ladle includes first and second heaters for heating the liquid metal, the first heater being located outside the filter medium and the second heater being an electrical immersion heater which is located in the inlet or outlet and adjacent the filter medium. The second heater is arranged for immersion in liquid metal flowing through the inlet or the outlet.