Isaac Padberg iaverypadberg

Building a background dataset from scratch

For object detection models with very few classes (<5) and no background images, most models will struggle with false positives. The model is only learning what the thing is that it needs to detect. This is only half the problem.

The other half is that the model needs to learn what not to detect. Heres where background images come in.

Aquire background images

The easiest way to get some background images is to take a video with a smart phone and chop it into individual frames.

	import org.jetbrains.kotlinx.multik.api.toNDArray
	import org.jetbrains.kotlinx.multik.ndarray.data.get
	import org.jetbrains.kotlinx.multik.ndarray.data.slice
	import org.jetbrains.kotlinx.multik.ndarray.operations.last
	import org.jetbrains.kotlinx.multik.ndarray.operations.minus
	import org.jetbrains.kotlinx.multik.ndarray.operations.plus
	import org.jetbrains.kotlinx.multik.ndarray.operations.toList

	// Filter out all items
	val startList = listOf(Pair(1,.2),Pair(2,.4),Pair(3,.6),Pair(4,.3))

	from sqlite3 import adapt
	import numpy as np



	from PIL import Image
	import glob
	import cv2
	import numpy as np
	import tflite_runtime.interpreter as tflite

	private fun argsort(needsSorting : Array<Int>){
	val theRange = IntRange(0,needsSorting.size-1)
	val idxs = theRange.toList().sortedBy { needsSorting[it] }
	}