Zhang Menghe ZhangMenghe

Using

Ubuntu 18.02
Android NDK 20.0.5594570
Protobuf v3.12.2
Language C++ / Java

Use clang:

S1: Generate standalone toolchain

Install MediaPipe

Follow the stepsHERE

Ubuntu 16.04

Can build the Brazel project for both desktop and Android:
- ubuntu $ bazel build -c opt \ --define MEDIAPIPE_DISABLE_GPU=1 \ --define no_aws_support=true \ mediapipe/examples/desktop/object_detection:object_detection_tensorflow and then `$ GLOG_logtostderr=1 bazel-bin/mediapipe/examples/desktop/object_detection/object_detection_tensorflow
  --calculator_graph_config_file=mediapipe/graphs/object_detection/object_detection_desktop_tensorflow_graph.pbtxt \

Tags: Unsupervised Learning, Monocular Video, KITTI

Unsupervised learning framework for task of monocular depth and camera motion estimation from unstructured video sequences.

Backgrounds

Warping-based view synthesis

create novel views of a specific subject from images fromdifferent view points
explicitly reconstruct the accurate 3D model.
forced to learn intermediate precictions of geometry and/or correspondences.
View Synthesis itself is mostly a graphics problem, and able to work in an end-to-end learning-based framework. However, in this way, geometry correspondences lose.
Related papers:

MASK R-CNN

Tags: 2017 Instance Segmentation

Priors

Faster R-CNN for object detection

Implementation

Steps

Scan the image and generate the proposals(areas likely to contain an object)
Classify the proposals and generate bounding boxes and masks.

Backbone

ResNet50/ResNet101: feature extractor(early layers for low-level features, later for high-level features.)

Basics

ls dir file listing -al - with hidden files
rm del -r to remove directory, -f = shift+del
cp srcFile dstDir copy file to dstDir with same name, -r cp directory
mv srcFile dstFile move or rename

System

alias see all alias
change alias or export permanatly

If you have a GPU, you shouldn't care about AVX support, because most expensive ops will be dispatched on a GPU device (unless explicitly set not to). In this case, you can simply ignore this warning by

# Just disables the warning, doesn't enable AVX/FMA
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

... or by setting export TF_CPP_MIN_LOG_LEVEL=2 if you're on Unix. Tensorflow is working fine anyway, but you won't see these annoying warnings.

	# Built application files
	/*/build/

	# Crashlytics configuations
	com_crashlytics_export_strings.xml

	# Local configuration file (sdk path, etc)
	local.properties

	# Gradle generated files

	from collections import deque # use to draw the last position line
	import numpy as np
	import argparse
	import imutils # helpful library (pyimagesearch.com)
	import cv2
	from centroidtracker import CentroidTracker
	import time
	# construct argument parse input
	# ap = argparse.ArgumentParser()
	# ap.add_argument("-v", "--video", help="path to the (optional) video file")

	//M1: exit and re-enter the application
	Intent mStartActivity = new Intent(this, MainActivity.class);
	int mPendingIntentId = 123456;
	PendingIntent mPendingIntent = PendingIntent.getActivity(this, mPendingIntentId, mStartActivity, PendingIntent.FLAG_CANCEL_CURRENT);
	AlarmManager mgr = (AlarmManager)this.getSystemService(Context.ALARM_SERVICE);
	mgr.set(AlarmManager.RTC, System.currentTimeMillis() + 100, mPendingIntent);
	System.exit(0);

	//M2: let the current activity finish, app itself won't close
	Intent i = new Intent(this, MainActivity.class);

	#!/usr/bin/env python
	import sys

	import tensorflow as tf
	import numpy as np

	from tensorflow.examples.tutorials.mnist import input_data

	class Model(object):
	def __init__(self, x, with_new_layer=False):