From 01a0801e5b04623170cf66980aeff0278ae446fd Mon Sep 17 00:00:00 2001
From: Wanming Lin <wanming.lin@intel.com>
Date: Thu, 2 Nov 2023 11:25:30 +0800
Subject: [PATCH] Update to align with latest spec

---
 image_classification/main.js               | 11 +++++++----
 image_classification/mobilenetv2_7_nchw.js | 11 ++++++-----
 2 files changed, 13 insertions(+), 9 deletions(-)

diff --git a/image_classification/main.js b/image_classification/main.js
index db5c228a..912cdca7 100644
--- a/image_classification/main.js
+++ b/image_classification/main.js
@@ -124,8 +124,9 @@ async function renderCamStream() {
   const inputCanvas = utils.getVideoFrame(camElement);
   console.log('- Computing... ');
   const start = performance.now();
-  await netInstance.compute(inputBuffer, outputBuffer);
+  const results = await netInstance.compute(inputBuffer, outputBuffer);
   computeTime = (performance.now() - start).toFixed(2);
+  outputBuffer = results.outputs.output;
   console.log(`  done in ${computeTime} ms.`);
   drawInput(inputCanvas, 'camInCanvas');
   showPerfResult();
@@ -256,7 +257,7 @@ async function main() {
       // UI shows model loading progress
       await ui.showProgressComponent('current', 'pending', 'pending');
       console.log('- Loading weights... ');
-      const contextOptions = {type: 'webnn', devicePreference: devicePreference};
+      const contextOptions = {type: 'webnn', deviceType: devicePreference};
       if (powerPreference) {
         contextOptions['powerPreference'] = powerPreference;
       }
@@ -281,11 +282,12 @@ async function main() {
       let medianComputeTime;
 
       // Do warm up
-      await netInstance.compute(inputBuffer, outputBuffer);
+      let results = await netInstance.compute(inputBuffer, outputBuffer);
 
       for (let i = 0; i < numRuns; i++) {
         start = performance.now();
-        await netInstance.compute(inputBuffer, outputBuffer);
+        results = await netInstance.compute(
+            results.inputs.input, results.outputs.output);
         computeTime = (performance.now() - start).toFixed(2);
         console.log(`  compute time ${i+1}: ${computeTime} ms`);
         computeTimeArray.push(Number(computeTime));
@@ -295,6 +297,7 @@ async function main() {
         medianComputeTime = medianComputeTime.toFixed(2);
         console.log(`  median compute time: ${medianComputeTime} ms`);
       }
+      outputBuffer = results.outputs.output;
       console.log('outputBuffer: ', outputBuffer);
       await ui.showProgressComponent('done', 'done', 'done');
       ui.readyShowResultComponents();
diff --git a/image_classification/mobilenetv2_7_nchw.js b/image_classification/mobilenetv2_7_nchw.js
index 016601bd..cb20a8ed 100644
--- a/image_classification/mobilenetv2_7_nchw.js
+++ b/image_classification/mobilenetv2_7_nchw.js
@@ -40,11 +40,11 @@ export class MobileNetV27Nchw {
     }
     if (clip) {
       // implement `clip` by `clamp` of  WebNN API
-      options.activation = this.builder_.clamp({minValue: 0, maxValue: 6});
+      return this.builder_.clamp(this.builder_.conv2d(input, weights, options), {minValue: 0, maxValue: 6});
     } else {
       options.activation = undefined;
+      return this.builder_.conv2d(input, weights, options);
     }
-    return this.builder_.conv2d(input, weights, options);
   }
 
   async buildGemm_(input, name) {
@@ -122,14 +122,14 @@ export class MobileNetV27Nchw {
     const conv3 = await this.buildConv_(bottleneck15, '95', true);
     const conv4 = await this.buildConv_(conv3, '97', false, {groups: 1280, strides: [7, 7]});
     const conv5 = await this.buildConv_(conv4, '104', false);
-    const reshape = this.builder_.reshape(conv5, [1, -1]);
+    const reshape = this.builder_.reshape(conv5, [1, null]);
     // return reshape;
     // const gemm = await this.buildGemm_(reshape, '104');
     return this.builder_.softmax(reshape);
   }
 
   async build(outputOperand) {
-    this.graph_ = await this.builder_.buildAsync({'output': outputOperand});
+    this.graph_ = await this.builder_.build({'output': outputOperand});
   }
 
   // Release the constant tensors of a model
@@ -143,6 +143,7 @@ export class MobileNetV27Nchw {
   async compute(inputBuffer, outputBuffer) {
     const inputs = {'input': inputBuffer};
     const outputs = {'output': outputBuffer};
-    await this.context_.compute(this.graph_, inputs, outputs);
+    const results = await this.context_.compute(this.graph_, inputs, outputs);
+    return results;
   }
 }