README

NEWS

Date	News	Version
Sept 2019	face recognition (insight face) was released for inferencing (STABLE), for training will available in the future version	>= v0.7.4
August 2019	deeplab semantic segmentation (PREVIEW version)was released	>= v0.6.10
Mei 2019	yolov3 object detection (STABLE version) was relesed	> v0.5.1
April 2019	Unet-segmentation (PREVIEW version)	v0.4.18

Tensorflow Compatibility

Tensorflow version	Simple-Tensor Version
1.4.1 - 1.1	<=v0.4.0
1.13.1 and 1.15.0	>=v0.4.3

ABOUT PROJECT

This project is a simplification of tensorflow operations and related projects

DEPENDENCIES

Tensorflow (1.4.1 - 1.15.0) pip3 install tensorflow-gpu
Comdutils pip3 install comdutils
opencv-python pip3 install opencv-python
numpy

HOW TO USE

:shipit: Installing The Package

pip3 install simple-tensor

:shipit: Import The Package

Tensor Operations

import tensorflow as tf
# tensor operations
from simple_tensor.tensor_operations import *
# tensor losses
from simple_tensor.tensor_losses import *
# tensor metrics
from simple_tensor.tensor_metrics import *

This packages contains tensor operations (conv2d, conv1d, depthwise conv2d, fully connected, conv2d transpose), tensor losses (softmax & sigmoid cross entropy, MSE), and tensor metrics (accuracy). For more detail documentations about tensor operations, visit this page

Transfer Learning Package (Image Classification)

import tensorflow as tf
from simple_tensor.transfer_learning.inception_utils import *
from simple_tensor.transfer_learning.inception_v4 import *

This package contains a library of tensorflow implementation of Inception-v4 for image classification. Densenet, Resnet, and VGG will be added in the future version. For more detail documentations about transfer learning package, visit this page

alt text

(img source: link)

Object Detector Package

import tensorflow as tf
from simple_tensor.object_detector.detector_utils import *
from simple_tensor.object_detector.yolo import Yolo

This package contains a library of tensorflow implementation of Yolov3 (training and inferencing). You can customize your yolo detector with four types of network ("big", 'medium", "small", "very_small"). For more detail documentations about object detector package (yolov3), visit this page.

alt text

(img source: pjreddie)

Unet Segmentation Package

import tensorflow as tf
from simple_tensor.segmentation.unet import UNet

This package contains the tensorflow implementation of U-net for semantic segmentation. For more detail, visit this page

alt text

(img source: internal)

Face Recognition Package (Insightface)

import tensorflow as tf
from simple_tensor.face_recog.insight_face import *

This package contains the tensorflow implementation of insight face. This repo is refractoring this github link. For more detail documentation, visit this page

LSTM Package

still on progress ....

Convert Keras Model to Tensorflow Serving

import tensorflow as tf
from simple_tensor.convert import *

DOCKER

We already prepared the all in one docker for computer vision and deep learning libraries, including tensorflow 1.12, Opencv3.4.2 and contrib, CUDA 9, CUDNN 7, Keras, jupyter, numpy, sklearn, scipy, statsmodel, pandas, matplotlib, seaborn, flask, gunicorn etc. See the list of dockerfile below:

Metadata

Release files for simple-tensor 1.1.6

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for simple-tensor 1.1.6
File	Size	Uploaded
simple_tensor-1.1.6.tar.gz	44.2 kB	Dec 2, 2022	Details

Release files / simple_tensor-1.1.6.tar.gz

Download URL	simple_tensor-1.1.6.tar.gz
Size	44.2 kB
Tags	Source
SHA-256 checksum How to use checksums	`9d71cbe99b72a2e78282c9092fffd92640fa2dad890462d7cda5711977a2b2f2`
BLAKE2b-256 checksum How to use checksums	`b521c823e3f80a2770d7e08bd3a151ad0dda21b4ca896b7a05fdc4bdca1ca605`
Upload date	Dec 2, 2022
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/4.0.1 CPython/3.8.10`