Please enable JavaScript.

Coggle requires JavaScript to display documents.

Sam's Carry On (Dataset (Load those datasets to disk (udacity dataset)…

- - - - Annotated Driving Dataset
        
        data 1
        
        car
        
        truck
        
        pedestrian
        
        data 2
        
        3 above + street light
- - - - caffe
        
        Link 2
        
        Link 1
- - - - matlab and caffe
    - - the tensorflow
    - - setup on aws ec2
        
        fast rcnn not the faster
        
        Korean guide
        
        github
        
        continue
        
        Continue on 7th
  - - - example
        
        udacity application
        
        parse the annotations option
        
        yolo v1 and yolo v2
        
        test against data2
        
        cfg provided
        
        trained w provided
        
        dependencies
        
        opencv 3
        
        python3
        
        numpy
        
        tensorflow 0.12
        
        trafic sign detection
        
        cfg file
        
        trained weight
        
        pre-compiled software
        
        collected data
    - - base system: done
      - cuda and openCV
  - - - obsclussion handling
    - - To realize road traffic flow
    - - udacity
        
        udacity - various environments which contain poor visibility conditions
    - - Haar Cascade
    - - Self-driving Cars — OpenCV and SVM Machine Learning with Scikit-Learn for Vehicle Detection on the Road
    - - tools
      - dataset
      - architecture : vonet
- - - - done
    - - guide
- - - - Keras
      - straight tensorflow
      - TF slim
      - TF learn
        github tutorial
    - - caffe2
        -github
        -tutorial
- - - - I have considered the problem as a specific instance of a sequence-to-sequence mapping problem:
        the model had to map the sequence of images to an equal-length sequence of the steering decisions
        
        The model architecture was composed of a convolutional vision stack followed by a stateful recurrent network.
        
        Well-established methods like residual connections, layer normalization and an aggressive regularization via dropout were necessary to obtain the best results.
        
        it allowed the model to build motion detectors and account for the driving dynamics. The recurrent cell tracked the state of the vehicle, which was supported by an auxiliary cost function — the model tried to predict not only the steering angle, but also the steering wheel torque and the speed of the vehicle.
    - - Our final model consisted of 3 streams that we merged at the final layer. Two of the streams were inspired by the NVIDIA self-driving car paper, and one of the streams was inspired by comma.ai’s steering model.