@tensorflow-models/body-pix
Version:
Pretrained BodyPix model in TensorFlow.js
236 lines • 12 kB
JavaScript
;
/**
* @license
* Copyright 2019 Google Inc. All Rights Reserved.
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
* =============================================================================
*/
var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
return new (P || (P = Promise))(function (resolve, reject) {
function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
step((generator = generator.apply(thisArg, _arguments || [])).next());
});
};
var __generator = (this && this.__generator) || function (thisArg, body) {
var _ = { label: 0, sent: function() { if (t[0] & 1) throw t[1]; return t[1]; }, trys: [], ops: [] }, f, y, t, g;
return g = { next: verb(0), "throw": verb(1), "return": verb(2) }, typeof Symbol === "function" && (g[Symbol.iterator] = function() { return this; }), g;
function verb(n) { return function (v) { return step([n, v]); }; }
function step(op) {
if (f) throw new TypeError("Generator is already executing.");
while (g && (g = 0, op[0] && (_ = 0)), _) try {
if (f = 1, y && (t = op[0] & 2 ? y["return"] : op[0] ? y["throw"] || ((t = y["return"]) && t.call(y), 0) : y.next) && !(t = t.call(y, op[1])).done) return t;
if (y = 0, t) op = [op[0] & 2, t.value];
switch (op[0]) {
case 0: case 1: t = op; break;
case 4: _.label++; return { value: op[1], done: false };
case 5: _.label++; y = op[1]; op = [0]; continue;
case 7: op = _.ops.pop(); _.trys.pop(); continue;
default:
if (!(t = _.trys, t = t.length > 0 && t[t.length - 1]) && (op[0] === 6 || op[0] === 2)) { _ = 0; continue; }
if (op[0] === 3 && (!t || (op[1] > t[0] && op[1] < t[3]))) { _.label = op[1]; break; }
if (op[0] === 6 && _.label < t[1]) { _.label = t[1]; t = op; break; }
if (t && _.label < t[2]) { _.label = t[2]; _.ops.push(op); break; }
if (t[2]) _.ops.pop();
_.trys.pop(); continue;
}
op = body.call(thisArg, _);
} catch (e) { op = [6, e]; y = 0; } finally { f = t = 0; }
if (op[0] & 5) throw op[1]; return { value: op[0] ? op[1] : void 0, done: true };
}
};
Object.defineProperty(exports, "__esModule", { value: true });
var tfconv = require("@tensorflow/tfjs-converter");
var tf = require("@tensorflow/tfjs-core");
// tslint:disable-next-line: no-imports-from-dist
var jasmine_util_1 = require("@tensorflow/tfjs-core/dist/jasmine_util");
var bodyPixModel = require("./body_pix_model");
var resnet = require("./resnet");
var util = require("./util");
(0, jasmine_util_1.describeWithFlags)('BodyPix', jasmine_util_1.ALL_ENVS, function () {
var bodyPix;
var inputResolution = 513;
var outputStride = 32;
var quantBytes = 4;
var numKeypoints = 17;
var numParts = 24;
var outputResolution = (inputResolution - 1) / outputStride + 1;
beforeAll(function () { return __awaiter(void 0, void 0, void 0, function () {
var resNetConfig;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
resNetConfig = { architecture: 'ResNet50', outputStride: outputStride, inputResolution: inputResolution, quantBytes: quantBytes };
spyOn(tfconv, 'loadGraphModel').and.callFake(function () {
return null;
});
spyOn(resnet, 'ResNet').and.callFake(function () {
return {
outputStride: outputStride,
predict: function (input) {
return {
inputResolution: inputResolution,
heatmapScores: tf.zeros([outputResolution, outputResolution, numKeypoints]),
offsets: tf.zeros([outputResolution, outputResolution, 2 * numKeypoints]),
displacementFwd: tf.zeros([outputResolution, outputResolution, 2 * (numKeypoints - 1)]),
displacementBwd: tf.zeros([outputResolution, outputResolution, 2 * (numKeypoints - 1)]),
segmentation: tf.zeros([outputResolution, outputResolution, 1]),
partHeatmaps: tf.zeros([outputResolution, outputResolution, numParts]),
longOffsets: tf.zeros([outputResolution, outputResolution, 2 * numKeypoints]),
partOffsets: tf.zeros([outputResolution, outputResolution, 2 * numParts])
};
},
dipose: function () { }
// tslint:disable-next-line:no-any
};
});
return [4 /*yield*/, bodyPixModel.load(resNetConfig)];
case 1:
bodyPix = _a.sent();
return [2 /*return*/];
}
});
}); });
it('segmentPerson does not leak memory', function () { return __awaiter(void 0, void 0, void 0, function () {
var input, beforeTensors;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
beforeTensors = tf.memory().numTensors;
return [4 /*yield*/, bodyPix.segmentPerson(input)];
case 1:
_a.sent();
expect(tf.memory().numTensors).toEqual(beforeTensors);
return [2 /*return*/];
}
});
}); });
it('segmentMultiPerson does not leak memory', function () { return __awaiter(void 0, void 0, void 0, function () {
var input, beforeTensors;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
beforeTensors = tf.memory().numTensors;
return [4 /*yield*/, bodyPix.segmentMultiPerson(input)];
case 1:
_a.sent();
expect(tf.memory().numTensors).toEqual(beforeTensors);
return [2 /*return*/];
}
});
}); });
it('segmentPersonParts does not leak memory', function () { return __awaiter(void 0, void 0, void 0, function () {
var input, beforeTensors;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
beforeTensors = tf.memory().numTensors;
return [4 /*yield*/, bodyPix.segmentPersonParts(input)];
case 1:
_a.sent();
expect(tf.memory().numTensors).toEqual(beforeTensors);
return [2 /*return*/];
}
});
}); });
it('segmentMultiPersonParts does not leak memory', function () { return __awaiter(void 0, void 0, void 0, function () {
var input, beforeTensors;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
beforeTensors = tf.memory().numTensors;
return [4 /*yield*/, bodyPix.segmentMultiPersonParts(input)];
case 1:
_a.sent();
expect(tf.memory().numTensors).toEqual(beforeTensors);
return [2 /*return*/];
}
});
}); });
it("segmentPerson uses default values when null is " +
"passed in inferenceConfig parameters", function () { return __awaiter(void 0, void 0, void 0, function () {
var input;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
spyOn(util, 'toInputResolutionHeightAndWidth').and.callThrough();
return [4 /*yield*/, bodyPix.segmentPerson(input, {})];
case 1:
_a.sent();
expect(util.toInputResolutionHeightAndWidth)
.toHaveBeenCalledWith('medium', 32, [73, 73]);
return [2 /*return*/];
}
});
}); });
it("segmentMultiPerson uses default values when null is " +
"passed in inferenceConfig parameters", function () { return __awaiter(void 0, void 0, void 0, function () {
var input;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
spyOn(util, 'toInputResolutionHeightAndWidth').and.callThrough();
return [4 /*yield*/, bodyPix.segmentMultiPerson(input, {})];
case 1:
_a.sent();
expect(util.toInputResolutionHeightAndWidth)
.toHaveBeenCalledWith('medium', 32, [73, 73]);
return [2 /*return*/];
}
});
}); });
it("segmentPersonParts uses default values when null is " +
"passed in inferenceConfig parameters", function () { return __awaiter(void 0, void 0, void 0, function () {
var input;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
spyOn(util, 'toInputResolutionHeightAndWidth').and.callThrough();
return [4 /*yield*/, bodyPix.segmentPersonParts(input, {})];
case 1:
_a.sent();
expect(util.toInputResolutionHeightAndWidth)
.toHaveBeenCalledWith('medium', 32, [73, 73]);
return [2 /*return*/];
}
});
}); });
it("segmentMultiPersonParts uses default values when null is " +
"passed in inferenceConfig parameters", function () { return __awaiter(void 0, void 0, void 0, function () {
var input;
return __generator(this, function (_a) {
switch (_a.label) {
case 0:
input = tf.zeros([73, 73, 3]);
spyOn(util, 'toInputResolutionHeightAndWidth').and.callThrough();
return [4 /*yield*/, bodyPix.segmentMultiPersonParts(input, {})];
case 1:
_a.sent();
expect(util.toInputResolutionHeightAndWidth)
.toHaveBeenCalledWith('medium', 32, [73, 73]);
return [2 /*return*/];
}
});
}); });
});
//# sourceMappingURL=body_pix_test.js.map