SegNet

SegNet is a TensorFlow implementation of the segmentation network proposed by Kendall et al., with cool features like strided deconvolution, a minified architecture and more.

Configuration

Create a config.py file, containing color maps, working dataset and other options.

autoencoder = 'segnet'
colors = {
  'segnet-32': [
    [64, 128, 64],   # Animal
    [192, 0, 128],   # Archway
    [0, 128, 192],   # Bicyclist
    [0, 128, 64],    # Bridge
    [128, 0, 0],     # Building
    [64, 0, 128],    # Car
    [64, 0, 192],    # CartLuggagePram
    [192, 128, 64],  # Child
    [192, 192, 128], # Column_Pole
    [64, 64, 128],   # Fence
    [128, 0, 192],   # LaneMkgsDriv
    [192, 0, 64],    # LaneMkgsNonDriv
    [128, 128, 64],  # Misc_Text
    [192, 0, 192],   # MotorcycleScooter
    [128, 64, 64],   # OtherMoving
    [64, 192, 128],  # ParkingBlock
    [64, 64, 0],     # Pedestrian
    [128, 64, 128],  # Road
    [128, 128, 192], # RoadShoulder
    [0, 0, 192],     # Sidewalk
    [192, 128, 128], # SignSymbol
    [128, 128, 128], # Sky
    [64, 128, 192],  # SUVPickupTruck
    [0, 0, 64],      # TrafficCone
    [0, 64, 64],     # TrafficLight
    [192, 64, 128],  # Train
    [128, 128, 0],   # Tree
    [192, 128, 192], # Truck_Bus
    [64, 0, 64],     # Tunnel
    [192, 192, 0],   # VegetationMisc
    [0, 0, 0],       # Void
    [64, 192, 0]     # Wall
  ]
}
gpu_memory_fraction = 0.7
strided = True
working_dataset = 'segnet-32'

Two kinds of architectures are supported at the moment: the original SegNet Encoder-Decoder (segnet), and a smaller version of the same (mini), which can be used for simpler segmentation problems. I suggest to use strided = True for faster and more reliable results.

The dataset_name needs to match the data directories you create in your input folder. You can use segnet-32 and segnet-13 to replicate the aforementioned Kendall et al. experiments.

Train and test

Generate your TFRecords using tfrecorder.py. In order to do so, put your PNG images in a raw folder, as follows:

input/
    raw/
        train/
        train-labels/
        test/
        test-labels/

Once you have your TFRecords, train SegNet with python src/train.py. Analogously, test it with python src/test.py.

SegNet-like Autoencoders in TensorFlow

Related tags

Overview

SegNet

Configuration

Train and test

Owner

Andrea Azzini

KwaiRec: A Fully-observed Dataset for Recommender Systems (Density: Almost 100%)

A whale detector design for the Kaggle whale-detector challenge!

Unofficial implement with paper SpeakerGAN: Speaker identification with conditional generative adversarial network

This package proposes simplified exporting pytorch models to ONNX and TensorRT, and also gives some base interface for model inference.

Official Code for AdvRush: Searching for Adversarially Robust Neural Architectures (ICCV '21)

A Unified Framework and Analysis for Structured Knowledge Grounding

Official implementation for paper: A Latent Transformer for Disentangled Face Editing in Images and Videos.

ScaleNet: A Shallow Architecture for Scale Estimation

The official implementation for ACL 2021 "Challenges in Information Seeking QA: Unanswerable Questions and Paragraph Retrieval".

3DIAS: 3D Shape Reconstruction with Implicit Algebraic Surfaces (ICCV 2021)

Generative Models as a Data Source for Multiview Representation Learning

Charsiu: A transformer-based phonetic aligner

Objective of the repository is to learn and build machine learning models using Pytorch. 30DaysofML Using Pytorch

Tensorflow implementation of Human-Level Control through Deep Reinforcement Learning

Controlling the MicriSpotAI robot from scratch

Beyond Image to Depth: Improving Depth Prediction using Echoes (CVPR 2021)

A library for answering questions using data you cannot see

Official Pytorch implementation of Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations

v objective diffusion inference code for JAX.

PyTorch implementation of the YOLO (You Only Look Once) v2