Home

Awesome

TLXCV

A Platform-agnostic Computer Vision Application Library, based on TensorLayerX.

Introduction

TLXCV provides a set of algorithms and high-level APIs for computer vision tasks, such as image classification, object detection, semantic segmentation, etc.
Some of the algorithms are converted from PaddlePaddle implementations, and some are implemented from scratch.

Quick Start

Installation

git clone https://github.com/tensorlayer/TLXCV.git
cd TLXCV
pip install -e .

train

python demo/image_classification/train.py

predict

python demo/image_classification/predict.py

模型列表 Models

分类模型 Classification

序号模型类别误差前后误差状态参考
1vgg16(pretrained model)一致0.0完成PaddleClas
2alexnet(pretrained model)一致0.0完成PaddleClas
3resnet50(pretrained model)一致0.0完成PaddleClas
4resnet101(pretrained model)一致0.0完成PaddleClas
5googlenet(pretrained model)一致0.0完成PaddleClas
6mobilenetv1(pretrained model)一致0.0完成PaddleClas
7mobilenetv2(pretrained model)一致0.0完成PaddleClas
8mobilenetv3(pretrained model)一致0.0完成PaddleClas
9shufflenetv2(pretrained model)一致0.0完成PaddleClas
10squeezenet(pretrained model)一致0.0完成PaddleClas
11inceptionv3(pretrained model)一致0.0完成PaddleClas
12regnet(pretrained model)一致0.0完成PaddleClas
13tnt(pretrained model)一致0.0完成PaddleClas
14darknet53(pretrained model)一致0.0完成PaddleClas
15densenet(pretrained model)一致0.0完成PaddleClas
16rednet50(pretrained model)一致0.0完成PaddleClas
17rednet101(pretrained model)一致0.0完成PaddleClas
18cspdarknet53(pretrained model)一致0.0完成PaddleClas
19efficientnet_b1(pretrained model)一致0.0完成PaddleClas
20efficientnet_b7(pretrained model)一致0.0完成PaddleClas
21dla34(pretrained model)一致0.0完成PaddleClas
22dla102(pretrained model)一致0.0完成PaddleClas
23dpn68(pretrained model)一致0.0完成PaddleClas
24dpn107(pretrained model)一致0.0完成PaddleClas
25ghostnet(pretrained model)一致0.0完成PaddleClas
26hardnet39(pretrained model)一致0.0完成PaddleClas
27hardnet85(pretrained model)一致0.0完成PaddleClas
28resnest50(pretrained model)一致0.0完成PaddleClas
29resnext50(pretrained model)一致0.0完成PaddleClas
30resnext101(pretrained model)一致0.0完成PaddleClas
31rexnet(pretrained model)微小误差0.00061244145完成PaddleClas
32se_resnext(pretrained model)一致0.0完成PaddleClas
33esnet_x0_5(pretrained model)一致0.0完成PaddleClas
34esnet_x1_0(pretrained model)一致0.0完成PaddleClas
35vit(pretrained model)一致0.0完成PaddleClas
36alt_gvt_small(pretrained model)一致0.0完成PaddleClas
37alt_gvt_base(pretrained model)一致0.0完成PaddleClas
38swin_transformer_base(pretrained model)0.0PaddleClas
39swin_transformer_small(pretrained model)0.0PaddleClas
40pcpvt_base(pretrained model)一致0.0完成PaddleClas
41pcpvt_large(pretrained model)一致0.0完成PaddleClas
42xception41(pretrained model)一致0.0完成PaddleClas
43xception65(pretrained model)一致0.0完成PaddleClas
44xception41_deeplab(pretrained model)一致0.0完成PaddleClas
45xception65_deeplab(pretrained model)一致0.0完成PaddleClas
46levit(pretrained model)一致0.0完成PaddleClas
47mixnet(pretrained model)微小误差0.00048300158完成PaddleClas
48convnext(pretrained model)一致0.0完成PaddleClas
49cswin(pretrained model)一致0.0完成PaddleClas
50deittiny(pretrained model)一致0.0完成PaddleClas
51deitsmall(pretrained model)一致0.0完成PaddleClas
52deitbase(pretrained model)一致0.0完成PaddleClas
53dvt(pretrained model)一致0.0完成PaddleClas
54peleenet(pretrained model)一致0.0完成PaddleClas
55pp_hgnet(pretrained model)一致0.0完成PaddleClas
56pp_lcnet(pretrained model)一致0.0完成PaddleClas
57pp_lcnet_v2(pretrained model)一致0.0完成PaddleClas
58pvt_v2(pretrained model)一致0.0完成PaddleClas
59res2net(pretrained model)一致0.0完成PaddleClas
60van(pretrained model)一致0.0完成PaddleClas

分割模型 Segmentation

序号模型前后误差状态参考
1fast_scnn0.0完成PaddleSeg
2hrnet0.0完成PaddleSeg
3encnet0.0完成PaddleSeg
4bisenet0.0完成PaddleSeg
5fastfcn0.0完成PaddleSeg
6enet0.0完成PaddleSeg

检测模型 Detection

序号模型前后误差状态方向
1yolov30.0完成PaddleDec
2ssd0.0完成PaddleDec
3yolox0.0完成PaddleDec
4picodet_lcnet0.0完成PaddleDec
5fcos_r500.0完成PaddleDec
6fcos_dcn0.0完成PaddleDec
7RetinaNet0.0完成PaddleDec
8Mask_RCNN0.0完成PaddleDec
9Faster_RCNN0.0完成PaddleDec
10CascadeRCNN0.0完成PaddleDec
11SOLOv20.0完成PaddleDec
12GFL0.0完成PaddleDec
13TOOD0.0完成PaddleDec
14CenterNet0.0完成PaddleDec
15TTFNet0.0完成PaddleDec

遥感模型 Remote Sensing

序号模型前后误差状态参考
1bit0.0完成PaddleRS
2cdnet0.0完成PaddleRS
3stanet0.0完成PaddleRS
4fcef0.0完成PaddleRS
5fccdn0.0完成PaddleRS
6dsamnet0.0完成PaddleRS
7snunet0.0完成PaddleRS
8dsifn0.0完成PaddleRS
9unet0.0完成PaddleRS
10farseg0.0完成PaddleRS
11deeplab0.0完成PaddleRS