运行segment anything模型的web demo 教程

这个web应用放在在源码的demo文件夹里：

在这里插入图片描述

这个前端仅基于React的web演示了如何加载固定图像和相应的SAM image embedding的.npy文件。

运行需要配置npm环境。

首先导出onnx的模型：

import torch
import numpy as np
import cv2
import matplotlib.pyplot as plt
from segment_anything import sam_model_registry, SamPredictor
from segment_anything.utils.onnx import SamOnnxModel

import onnxruntime
from onnxruntime.quantization import QuantType
from onnxruntime.quantization.quantize import quantize_dynamic

# 我本地存在checkpoints/sam_vit_h_4b8939.pth
checkpoint =  "../checkpoints/sam_vit_h_4b8939.pth"
model_type = "vit_h"

sam = sam_model_registry[model_type](checkpoint=checkpoint)

onnx_model_path = None  # Set to use an already exported model, then skip to the next section.

import warnings

onnx_model_path = "sam_onnx_example.onnx"

onnx_model = SamOnnxModel(sam, return_single_mask=True)

dynamic_axes = {
    "point_coords": {1: "num_points"},
    "point_labels": {1: "num_points"},
}

embed_dim = sam.prompt_encoder.embed_dim
embed_size = sam.prompt_encoder.image_embedding_size
mask_input_size = [4 * x for x in embed_size]
dummy_inputs = {
    "image_embeddings": torch.randn(1, embed_dim, *embed_size, dtype=torch.float),
    "point_coords": torch.randint(low=0, high=1024, size=(1, 5, 2), dtype=torch.float),
    "point_labels": torch.randint(low=0, high=4, size=(1, 5), dtype=torch.float),
    "mask_input": torch.randn(1, 1, *mask_input_size, dtype=torch.float),
    "has_mask_input": torch.tensor([1], dtype=torch.float),
    "orig_im_size": torch.tensor([1500, 2250], dtype=torch.float),
}
output_names = ["masks", "iou_predictions", "low_res_masks"]

with warnings.catch_warnings():
    warnings.filterwarnings("ignore", category=torch.jit.TracerWarning)
    warnings.filterwarnings("ignore", category=UserWarning)
    # 这里导出了sam_onnx_example.onnx
    with open(onnx_model_path, "wb") as f:
        torch.onnx.export(
            onnx_model,
            tuple(dummy_inputs.values()),
            f,
            export_params=True,
            verbose=False,
            opset_version=17,
            do_constant_folding=True,
            input_names=list(dummy_inputs.keys()),
            output_names=output_names,
            dynamic_axes=dynamic_axes,
        )   
        
        
onnx_model_quantized_path = "sam_onnx_quantized_example.onnx"
quantize_dynamic(
    model_input=onnx_model_path,
    model_output=onnx_model_quantized_path,
    # 这个实际运行的时候，会报错
    #optimize_model=True,
    per_channel=False,
    reduce_range=False,
    weight_type=QuantType.QUInt8,
)
onnx_model_path = onnx_model_quantized_path

这样会生成两个onnx模型：

sam_onnx_example.onnx
sam_onnx_example.onnx 这个模型是所需的，需要

然后选一个示例图像dog.jpg进行编码，输出其.npy的编码文件：

# 注意，这个一定要重新导入，因为下面的代码使用的cuda加速的，上面导出模型用的是CPU模式
checkpoint = "../checkpoints/sam_vit_h_4b8939.pth"
model_type = "vit_h"
sam = sam_model_registry[model_type](checkpoint=checkpoint)
sam.to(device='cuda')
predictor = SamPredictor(sam)

image = cv2.imread('../demo/src/assets/data/dogs.jpg')
predictor.set_image(image)
image_embedding = predictor.get_image_embedding().cpu().numpy()
np.save("dogs_embedding.npy", image_embedding)
type(image_embedding),image_embedding.shape

按照demo/src/App.tsx规定的路径放置文件：

const IMAGE_PATH = "/assets/data/dogs.jpg";
const IMAGE_EMBEDDING = "/assets/data/dogs_embedding.npy";
const MODEL_DIR = "/model/sam_onnx_quantized_example.onnx";

运行demo：