Aloception’s documentation

Geting Started

• Getting started
  • About Aloception
  • Install
  • Other optional installation
• Aloscene: Computer vision with ease
  • Hello world
  • Exploring Frames
  • Playing with Frames
• Alodataset: Loading your vision datasets
  • Using samples
  • Stream loader
  • Train loader
  • Using transformations
• Alonet : Training your models
  • Datasets
  • Data modules
  • Training pipeline
• About augmented tensors
  • Statefull tensors
  • Augmented tensors and attached labels.
  • Transforming augmented tensors

Turorials

• How to setup your data?
  • 1. Prepare your data
  • 2. Setup a custom dataset based on COCO JSON
  • 3. Make a config file
• Training Detr
  • 1. LitDetr argument levels
  • 2. Train process
  • 3. Make inferences
• Finetuning DETR
  • 1. Train DETR50 Finetune
  • 2. Make inferences
• Training Panoptic Head module
  • 1. LitPanopticDetr and CocoPanoptic2Detr
  • 2. Train process
  • 3. Make inferences
• Training Deformable DETR
  • 1. Deformable DETR trainer
  • 2. Make inferences
  • 3. Model performance comparison
• Finetuning Deformanble DETR
  • 1. Train Deformable DETR50 Finetune
  • 2. Make inferences
  • 3. Optional: Make prediction in camera
• Exporting DETR / Deformable-DETR to TensorRT
  • DETR
  • Deformable DETR

Aloception API

• Aloscene
  • Frame
  • BoundingBoxes2D
  • OrientedBoxes2D
    • Basic usage
    • API
  • Points2D
  • Label
  • Mask
    • Basic Use
  • Mask API
  • Flow
    • Basic Use
    • Flow API
  • Disparity
    • Basic Use
    • Disparity API
  • BoundingBoxes3D
    • Visualization
    • Transformation
    • IoU / GIoU 3D
    • Other features
    • API
  • Camera calibration
  • Augmented Tensor
  • Spatial Augmented Tensor
• Alodataset
  • Transformations
  • Base dataset
  • Merging datasets
    • Basic usage
    • MergeDataset API
  • ChairsSDHom
    • ChairsSDHomDataset API
  • Coco detection
    • Basic use
    • Change of default database
    • Class filtering
    • Download a minimal dataset sample
    • Coco detection API
  • Crowd human
  • Flyingthings3D Subset
    • Flyingthings3DSubsetDataset API
  • FlyingChairs2
    • FlyingChairs2 API
  • Waymo Dataset
    • Dataset preparation
    • Basic usage
    • API
  • MOT17
  • Sintel Optical Flow
    • SintelFlowDataset API
  • Sintel Disparity
    • SintelDisparityDataset API
  • Sintel Multi (Flow + Disparity)
• Alonet
  • Detr
    • Models
    • Architecture
    • Training
    • Detr: Data connectors
  • Deformable DETR
    • Models
    • Architecture
    • Training
  • Detr panoptic
    • Models
    • Training
  • Callbacks
    • Base Metrics Callback
    • AP Metrics Callback
    • PQ Metrics Callback
    • Metrics Callbacks
    • Object Detector Callback
  • Metrics
    • AP Metrics
    • PQ Metrics
  • Transformer
    • MLP
    • Positional Encoding
    • Transformer