NEGEMMLowpMatrixMultiplyCore: fp16 output support for dequantization

[alvoron]

Feature request: NEGEMMLowpMatrixMultiplyCore fp16 output support for dequantization.

cc @milpuz01

[morgolock]

Hi @alvoron

Just to clarify, can you please confirm that you would like to be able to use FP16 as output in NEGEMMLowpMatrixMultiplyCore.h

~NEGEMMLowpMatrixMultiplyCore();
    /** Initialise the kernel's inputs, output
     *
     * Valid data layouts:
     * - NHWC
     * - NCHW
     *
     * Valid data type configurations:
     * |src0           |src1               |src2     |dst            |
     * |:--------------|:------------------|:--------|:--------------|
     * |QASYMM8        |QASYMM8            |S32      |QASYMM8        |
     * |QASYMM8        |QASYMM8_SIGNED     |S32      |QASYMM8        |
     * |QASYMM8        |QSYMM8_PER_CHANNEL |S32      |QASYMM8        |
     * |QASYMM8        |QSYMM8             |S32      |QASYMM8        |
     * |QASYMM8        |QASYMM8            |S32      |S32            |
     * |QASYMM8        |QSYMM8_PER_CHANNEL |S32      |S32            |
     * |QASYMM8        |QSYMM8             |S32      |S32            |
     * |QASYMM8        |QASYMM8_SIGNED     |F32      |F32            |
     * |QASYMM8_SIGNED |QASYMM8_SIGNED     |S32      |QASYMM8_SIGNED |
     * |QASYMM8_SIGNED |QSYMM8_PER_CHANNEL |S32      |QASYMM8_SIGNED |
     * |QASYMM8_SIGNED |QSYMM8             |S32      |QASYMM8_SIGNED |
     * |QASYMM8_SIGNED |QASYMM8_SIGNED     |S32      |S32            |
     * |QASYMM8_SIGNED |QSYMM8_PER_CHANNEL |S32      |S32            |
     * |QASYMM8_SIGNED |QSYMM8             |S32      |S32            |
     * |QASYMM8_SIGNED |QASYMM8_SIGNED     |F32      |F32            |