RayBNN_Cell

使用 CUDA、OpenCL 和 oneAPI 通过 GPU、CPU 和 FPGA 为 RayBNN 生成的单元位置生成库

安装 Arrayfire

在 https://arrayfire.com/binaries/ 安装 Arrayfire 3.9.0 二进制文件

或从源码构建 https://github.com/arrayfire/arrayfire/wiki/Getting-ArrayFire

添加到 Cargo.toml

arrayfire = { version = "3.8.1", package = "arrayfire_fork" }
rayon = "1.7.0"
num = "0.4.1"
num-traits = "0.2.16"
half = { version = "2.3.1" , features = ["num-traits"] }
RayBNN_DataLoader = "2.0.3"
RayBNN_Cell = "2.0.3"

示例列表

生成单元并检查碰撞

//Generate Random Uniform Cells within a Sphere
let mut cell_pos: arrayfire::Array<f32>  = RayBNN_Cell::Hidden::Sphere::generate_uniform_sphere_position(&modeldata_float, &modeldata_int);


//Get indicies of non colliding cells
let idx = RayBNN_Cell::Hidden::Sphere::check_cell_collision_minibatch(
    &modeldata_float, 
    &cell_pos
);

let idx = arrayfire::locate(&idx);

//Select non colliding cells
cell_pos = arrayfire::lookup(&cell_pos, &idx, 0);

单元碰撞检查器类型

//Checks One by one
let idx = RayBNN_Cell::Hidden::Sphere::check_cell_collision_serial(
    &modeldata_float, 
    &cell_pos
);

//Checks All cells at once
let idx = RayBNN_Cell::Hidden::Sphere::check_cell_collision_batch(
    &modeldata_float, 
    &cell_pos
);

//Checks in minibatches
let idx = RayBNN_Cell::Hidden::Sphere::check_cell_collision_minibatch(
    &modeldata_float, 
    &cell_pos
);

生成输入神经元位置

let input_neurons: arrayfire::Array<f64> = RayBNN_Cell::Input::Sphere::create_spaced_neurons_1D(
    sphere_rad,
    input_size,
);