BiSeNet based on pytorch

Last update: Dec 26, 2022

Related tags

Deep Learning BiSeNet

Overview

BiSeNet

BiSeNet based on pytorch 0.4.1 and python 3.6

Dataset

Download CamVid dataset from Google Drive or Baidu Yun(6xw4).

Pretrained model

Download best_dice_loss_miou_0.655.pth in Google Drive or in Baidu Yun(6y3e) and put it in ./checkpoints

Demo

python demo.py

Result

Original	GT	Predict

Train

python train.py

Use tensorboard to see the real-time loss and accuracy

loss on train

pixel precision on val

miou on val

Test

python test.py

Result

class	Bicyclist	Building	Car	Pole	Fence	Pedestrian	Road	Sidewalk	SignSymbol	Sky	Tree	miou
iou	0.61	0.80	0.86	0.35	0.37	0.59	0.88	0.81	0.28	0.91	0.73	0.655

This time I train the model with dice loss and get better result than cross entropy loss. I did not use lots special training strategy, you can get much better result than this repo if using task-specific strategy.
This repo is mainly for proving the effeciveness of the model.
I also tried some simplified version of bisenet but it seems does not preform very well in CamVid dataset.

Speed

Method	640×320	1280×720	1920×1080
Paper	129.4	47.9	23
This Repo	126.8	53.7	23.6

This shows the speed comparison between paper and my implementation.

The number in first row means input image resolution.
The number in second and third row means FPS.
The result is based on resnet-18.

Future work

Finish real-time segmentation with camera or pre-load video

BiSeNet based on pytorch

Related tags

Overview

BiSeNet

Dataset

Pretrained model

Demo

Result

Train

loss on train

pixel precision on val

miou on val

Test

Result

Speed

Future work

Reference

Owner

Face Mask Detection System built with OpenCV, TensorFlow using Computer Vision concepts

Supplementary code for SIGGRAPH 2021 paper: Discovering Diverse Athletic Jumping Strategies

Vector Quantized Diffusion Model for Text-to-Image Synthesis

custom pytorch implementation of MoCo v3

Code for the RA-L (ICRA) 2021 paper "SeqNet: Learning Descriptors for Sequence-Based Hierarchical Place Recognition"

Codes for our IJCAI21 paper: Dialogue Discourse-Aware Graph Model and Data Augmentation for Meeting Summarization

PocketNet: Extreme Lightweight Face Recognition Network using Neural Architecture Search and Multi-Step Knowledge Distillation

A script helps the user to update Linux and Mac systems through the terminal

🔮 Execution time predictions for deep neural network training iterations across different GPUs.

Source code for our paper "Empathetic Response Generation with State Management"

U^2-Net - Portrait matting This repository explores possibilities of using the original u^2-net model for portrait matting.

Implementation for ACProp ( Momentum centering and asynchronous update for adaptive gradient methdos, NeurIPS 2021)

Official PyTorch implementation of "Camera Distance-aware Top-down Approach for 3D Multi-person Pose Estimation from a Single RGB Image", ICCV 2019

Source code for paper "Deep Superpixel-based Network for Blind Image Quality Assessment"

Speckle-free Holography with Partially Coherent Light Sources and Camera-in-the-loop Calibration

Automatic Image Background Subtraction

[NeurIPS 2021] Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data

Interactive Image Generation via Generative Adversarial Networks

Kalman Filter book using Jupyter Notebook. Focuses on building intuition and experience, not formal proofs. Includes Kalman filters,extended Kalman filters, unscented Kalman filters, particle filters, and more. All exercises include solutions.

Spatial Sparse Convolution Library