creek

commit 8f5b90b67fcd6cc2d51f8f52d9fbccd3b514755c
parent 2cecc5ea83303efc4d86fb0294f37a49f352a2ea
Author: Sanjeebani Parida <sanjeebani14@gmail.com>
Date:   Thu, 27 Nov 2025 11:02:26 +0530

Fixed emotion and lighting model

Diffstat:

8 files changed, 242 insertions(+), 181 deletions(-)
diff --git a/lib/services/emotional_embeddings_service.dart b/lib/services/emotional_embeddings_service.dart
@@ -1,11 +1,9 @@
-// lib/services/style_analyzer_service.dart
-
 import 'dart:convert';
 import 'dart:math';
 import 'dart:io';
 import 'package:flutter/foundation.dart';
 import 'package:flutter/services.dart';
-import 'package:onnxruntime/onnxruntime.dart';
+import 'package:flutter_onnxruntime/flutter_onnxruntime.dart';
 import 'package:path_provider/path_provider.dart';
 import '../utils/clip_image_processor.dart';
 
@@ -17,84 +15,89 @@ class EmotionalEmbeddingsService {
   static Future<void> initialize() async {
     if (_session != null) return;
 
-    debugPrint("Initializing AI...");
-
-    // 1. Setup ONNX Environment
-    OrtEnv.instance.init();
-
-    // 2. Load Model from Assets
-    final sessionOptions = OrtSessionOptions();
-    final rawAsset = await rootBundle.load('assets/clip_vitb32_image.onnx');
-    final bytes = rawAsset.buffer.asUint8List();
+    try {
+      debugPrint("Initializing Embeddings (Dino/CLIP)...");
 
-    // Store temp file to load into ONNX
-    final dir = await getTemporaryDirectory();
-    final modelFile = File('${dir.path}/clip_model.onnx');
-    await modelFile.writeAsBytes(bytes);
+      final rawAsset = await rootBundle.load('assets/clip_image_encoder.onnx');
+      final dir = await getTemporaryDirectory();
+      final modelFile = File('${dir.path}/clip_model.onnx');
 
-    _session = OrtSession.fromFile(modelFile, sessionOptions);
+      if (!await modelFile.exists()) {
+        await modelFile.writeAsBytes(rawAsset.buffer.asUint8List());
+      }
 
-    // 3. Load Embeddings JSON
-    final jsonString = await rootBundle.loadString(
-      'assets/emotion_centroids.json',
-    );
-    final jsonData = json.decode(jsonString);
+      // Initialize runtime
+      final ort = OnnxRuntime();
+      _session = await ort.createSession(modelFile.path);
 
-    _classes = List<String>.from(jsonData['classes']);
-    _centroids =
-        (jsonData['centroids'] as List)
-            .map((e) => List<double>.from(e))
-            .toList();
+      final jsonString = await rootBundle.loadString(
+        'assets/emotion_centroids.json',
+      );
+      final jsonData = json.decode(jsonString);
 
-    debugPrint("Loaded ${_classes!.length} emotions.");
+      _classes = List<String>.from(jsonData['classes']);
+      _centroids =
+          (jsonData['centroids'] as List)
+              .map((e) => List<double>.from(e))
+              .toList();
+    } catch (e) {
+      debugPrint("Embeddings Init Error: $e");
+    }
   }
 
-  static Future<Map<String, dynamic>?> analyzeImage(String imagePath) async {
+  static Future<Map<String, dynamic>?> analyze(String imagePath) async {
     if (_session == null) await initialize();
 
+    OrtValue? inputOrt;
+    OrtValue? outputOrt;
+
     try {
-      // 1. Preprocess Image
       final float32Input = await ClipImageProcessor.preprocess(imagePath);
+      //debugPrint("[EMOTION] Preprocess length: ${float32Input?.length}");
       if (float32Input == null) return null;
 
-      // 2. Create Tensor [1, 3, 224, 224]
-      final inputTensor = OrtValueTensor.createTensorWithDataList(
-        float32Input,
-        [1, 3, 224, 224],
-      );
+      // Create Tensor
+      // Note: ensure ClipImageProcessor returns a flat List<double>
+      inputOrt = await OrtValue.fromList(float32Input, [1, 3, 224, 224]);
+
+      // Run Inference
+      // 'image' is the input name for CLIP. Check your model if it differs (e.g. 'input')
+      final outputs = await _session!.run({"image": inputOrt});
+      //debugPrint("[EMOTION] Output keys: ${outputs.keys}");
+
+      // Get Output
+      outputOrt =
+          outputs['output']; // or outputs['image_features'] depending on model
+      // If outputOrt is null, try iterating outputs.values.first
+      final resultOrt = outputOrt ?? outputs.values.first;
 
-      // 3. Run Inference
-      final inputs = {
-        'image': inputTensor,
-      }; // Check input name in Netron.app if 'image' fails (might be 'input')
-      final runOptions = OrtRunOptions();
-      final outputs = _session!.run(runOptions, inputs);
+      if (resultOrt == null) throw Exception("No output from model");
 
-      // 4. Get Embedding [1, 512]
-      // ONNX Runtime usually returns a List<List<double>> for 2D output
-      final rawOutput = outputs[0]?.value as List;
-      // Flatten if necessary, or access [0]
-      List<double> imgFeat =
-          (rawOutput[0] as List).map((e) => e as double).toList();
+      // Flatten Output
+      final List<dynamic> rawList = await resultOrt.asList();
+      final List<double> imgFeat = [];
+
+      void flatten(dynamic data) {
+        if (data is num)
+          imgFeat.add(data.toDouble());
+        else if (data is List)
+          for (var item in data) flatten(item);
+      }
 
-      // Cleanup
-      inputTensor.release();
-      runOptions.release();
-      outputs.forEach((element) => element?.release());
+      flatten(rawList);
 
-      // 5. Math: Normalize & Dot Product
-      imgFeat = _l2Normalize(imgFeat);
+      // debugPrint("[EMOTION] Embedding length: ${imgFeat.length}");
+      // debugPrint("[EMOTION] Centroid length: ${_centroids?[0].length}");
 
+      // Normalize & Compare
+      final normFeat = _l2Normalize(imgFeat);
       List<Map<String, dynamic>> scores = [];
+
       for (int i = 0; i < _classes!.length; i++) {
-        double score = _dotProduct(imgFeat, _centroids![i]);
-        scores.add({
-          "name": _classes![i],
-          "score": score * 100.0, // Scale for display
-        });
+        double score = _dotProduct(normFeat, _centroids![i]);
+        scores.add({"name": _classes![i], "score": score * 100.0});
       }
 
-      // 6. Sort & Return
       scores.sort(
         (a, b) => (b['score'] as double).compareTo(a['score'] as double),
       );
@@ -103,28 +106,28 @@ class EmotionalEmbeddingsService {
         "success": true,
         "label": scores.first['name'],
         "top5": scores.take(5).toList(),
-        "scores": {for (var e in scores) e['name']: e['score']},
       };
     } catch (e) {
-      debugPrint("Analysis Error: $e");
+      debugPrint("Embeddings Analysis Error: $e");
       return null;
+    } finally {
+      inputOrt?.dispose();
+      outputOrt?.dispose();
     }
   }
 
-  // --- MATH HELPERS ---
-
   static List<double> _l2Normalize(List<double> vec) {
-    double sum = 0;
-    for (var v in vec) {
-      sum += v * v;
-    }
+    double sum = vec.fold(0, (p, c) => p + c * c);
     double norm = sqrt(sum);
+    if (norm == 0) return vec;
     return vec.map((v) => v / norm).toList();
   }
 
   static double _dotProduct(List<double> a, List<double> b) {
     double sum = 0;
-    for (int i = 0; i < a.length; i++) {
+    // Ensure lengths match to avoid range errors
+    int len = min(a.length, b.length);
+    for (int i = 0; i < len; i++) {
       sum += a[i] * b[i];
     }
     return sum;
diff --git a/lib/services/image_analyzer_service.dart b/lib/services/image_analyzer_service.dart
@@ -15,7 +15,7 @@ class ImageAnalyzerService {
     // 1. Instantiate services
     final layoutService = LayoutAnalyzerService();
     final colorService = ColorAnalyzerService();
-    final textureService = TextureAnalyzerService();
+    final textureService = TextureAnalyzerService();    
     // EmbeddingService is static
 
     try {
@@ -24,7 +24,9 @@ class ImageAnalyzerService {
         layoutService.analyze(imagePath),           // Index 0: Layout
         colorService.analyze(imagePath),            // Index 1: Color
         textureService.analyze(imagePath),          // Index 2: Texture
-        EmbeddingAnalyzerService.analyze(imagePath) // Index 3: Embeddings
+        EmbeddingAnalyzerService.analyze(imagePath), // Index 3: Embeddings
+        EmotionalEmbeddingsService.analyze(imagePath), // Index 4: Emotional Embeddings
+        LightingEmbeddingsService.analyze(imagePath), // Index 5: Lighting Embeddings
       ]);
 
       // 3. Construct Unified Result
@@ -35,6 +37,8 @@ class ImageAnalyzerService {
         'color': results[1],
         'texture': results[2],   // This is a List
         'embedding': results[3],
+        'emotions': results[4],
+        'lighting': results[5],
       };
 
       return combinedResult;
diff --git a/lib/services/lighting_service.dart b/lib/services/lighting_service.dart
@@ -1,11 +1,9 @@
-// lib/services/style_analyzer_service.dart
-
 import 'dart:convert';
 import 'dart:math';
 import 'dart:io';
 import 'package:flutter/foundation.dart';
 import 'package:flutter/services.dart';
-import 'package:onnxruntime/onnxruntime.dart';
+import 'package:flutter_onnxruntime/flutter_onnxruntime.dart';
 import 'package:path_provider/path_provider.dart';
 import '../utils/clip_image_processor.dart';
 
@@ -17,84 +15,85 @@ class LightingEmbeddingsService {
   static Future<void> initialize() async {
     if (_session != null) return;
 
-    debugPrint("Initializing AI...");
-
-    // 1. Setup ONNX Environment
-    OrtEnv.instance.init();
-
-    // 2. Load Model from Assets
-    final sessionOptions = OrtSessionOptions();
-    final rawAsset = await rootBundle.load('assets/clip_vitb32_image.onnx');
-    final bytes = rawAsset.buffer.asUint8List();
+    try {
+      debugPrint("Initializing Embeddings (Dino/CLIP)...");
 
-    // Store temp file to load into ONNX
-    final dir = await getTemporaryDirectory();
-    final modelFile = File('${dir.path}/clip_model.onnx');
-    await modelFile.writeAsBytes(bytes);
+      final rawAsset = await rootBundle.load('assets/clip_image_encoder.onnx');
+      final dir = await getTemporaryDirectory();
+      final modelFile = File('${dir.path}/clip_model.onnx');
 
-    _session = OrtSession.fromFile(modelFile, sessionOptions);
+      if (!await modelFile.exists()) {
+        await modelFile.writeAsBytes(rawAsset.buffer.asUint8List());
+      }
 
-    // 3. Load Embeddings JSON
-    final jsonString = await rootBundle.loadString(
-      'assets/lighting_centroids.json',
-    );
-    final jsonData = json.decode(jsonString);
+      // Initialize runtime
+      final ort = OnnxRuntime();
+      _session = await ort.createSession(modelFile.path);
 
-    _classes = List<String>.from(jsonData['classes']);
-    _centroids =
-        (jsonData['centroids'] as List)
-            .map((e) => List<double>.from(e))
-            .toList();
+      final jsonString = await rootBundle.loadString(
+        'assets/lighting_centroids.json',
+      );
+      final jsonData = json.decode(jsonString);
 
-    debugPrint("Loaded ${_classes!.length} lighting styles.");
+      _classes = List<String>.from(jsonData['classes']);
+      _centroids =
+          (jsonData['centroids'] as List)
+              .map((e) => List<double>.from(e))
+              .toList();
+    } catch (e) {
+      debugPrint("Embeddings Init Error: $e");
+    }
   }
 
-  static Future<Map<String, dynamic>?> analyzeImage(String imagePath) async {
+  static Future<Map<String, dynamic>?> analyze(String imagePath) async {
     if (_session == null) await initialize();
 
+    OrtValue? inputOrt;
+    OrtValue? outputOrt;
+
     try {
-      // 1. Preprocess Image
       final float32Input = await ClipImageProcessor.preprocess(imagePath);
       if (float32Input == null) return null;
 
-      // 2. Create Tensor [1, 3, 224, 224]
-      final inputTensor = OrtValueTensor.createTensorWithDataList(
-        float32Input,
-        [1, 3, 224, 224],
-      );
+      // Create Tensor
+      // Note: ensure ClipImageProcessor returns a flat List<double>
+      inputOrt = await OrtValue.fromList(float32Input, [1, 3, 224, 224]);
+
+      // Run Inference
+      // 'image' is the input name for CLIP. Check your model if it differs (e.g. 'input')
+      final outputs = await _session!.run({"image": inputOrt});
+      //debugPrint("[LIGHTING] Output keys: ${outputs.keys}");
 
-      // 3. Run Inference
-      final inputs = {
-        'image': inputTensor,
-      }; // Check input name in Netron.app if 'image' fails (might be 'input')
-      final runOptions = OrtRunOptions();
-      final outputs = _session!.run(runOptions, inputs);
+      // Get Output
+      outputOrt =
+          outputs['output']; // or outputs['image_features'] depending on model
+      // If outputOrt is null, try iterating outputs.values.first
+      final resultOrt = outputOrt ?? outputs.values.first;
 
-      // 4. Get Embedding [1, 512]
-      // ONNX Runtime usually returns a List<List<double>> for 2D output
-      final rawOutput = outputs[0]?.value as List;
-      // Flatten if necessary, or access [0]
-      List<double> imgFeat =
-          (rawOutput[0] as List).map((e) => e as double).toList();
+      if (resultOrt == null) throw Exception("No output from model");
 
-      // Cleanup
-      inputTensor.release();
-      runOptions.release();
-      outputs.forEach((element) => element?.release());
+      // Flatten Output
+      final List<dynamic> rawList = await resultOrt.asList();
+      final List<double> imgFeat = [];
+
+      void flatten(dynamic data) {
+        if (data is num)
+          imgFeat.add(data.toDouble());
+        else if (data is List)
+          for (var item in data) flatten(item);
+      }
 
-      // 5. Math: Normalize & Dot Product
-      imgFeat = _l2Normalize(imgFeat);
+      flatten(rawList);
 
+      // Normalize & Compare
+      final normFeat = _l2Normalize(imgFeat);
       List<Map<String, dynamic>> scores = [];
+
       for (int i = 0; i < _classes!.length; i++) {
-        double score = _dotProduct(imgFeat, _centroids![i]);
-        scores.add({
-          "name": _classes![i],
-          "score": score * 100.0, // Scale for display
-        });
+        double score = _dotProduct(normFeat, _centroids![i]);
+        scores.add({"name": _classes![i], "score": score * 100.0});
       }
 
-      // 6. Sort & Return
       scores.sort(
         (a, b) => (b['score'] as double).compareTo(a['score'] as double),
       );
@@ -103,28 +102,28 @@ class LightingEmbeddingsService {
         "success": true,
         "label": scores.first['name'],
         "top5": scores.take(5).toList(),
-        "scores": {for (var e in scores) e['name']: e['score']},
       };
     } catch (e) {
-      debugPrint("Analysis Error: $e");
+      debugPrint("Embeddings Analysis Error: $e");
       return null;
+    } finally {
+      inputOrt?.dispose();
+      outputOrt?.dispose();
     }
   }
 
-  // --- MATH HELPERS ---
-
   static List<double> _l2Normalize(List<double> vec) {
-    double sum = 0;
-    for (var v in vec) {
-      sum += v * v;
-    }
+    double sum = vec.fold(0, (p, c) => p + c * c);
     double norm = sqrt(sum);
+    if (norm == 0) return vec;
     return vec.map((v) => v / norm).toList();
   }
 
   static double _dotProduct(List<double> a, List<double> b) {
     double sum = 0;
-    for (int i = 0; i < a.length; i++) {
+    // Ensure lengths match to avoid range errors
+    int len = min(a.length, b.length);
+    for (int i = 0; i < len; i++) {
       sum += a[i] * b[i];
     }
     return sum;
diff --git a/lib/ui/widgets/analysis_dialog.dart b/lib/ui/widgets/analysis_dialog.dart
@@ -99,6 +99,87 @@ class AnalysisDialog extends StatelessWidget {
                         )).toList(),
                       ),
                     ),
+                  
+                  // 5. Emotional Embeddings
+                  // 5. Emotional Embeddings
+                  if (data['emotions'] != null)
+                    _buildSection(
+                      "Emotional Tone",
+                      Icons.mood,
+                      Colors.redAccent,
+                      Column(
+                        crossAxisAlignment: CrossAxisAlignment.start,
+                        children: [
+
+                          Text(
+                            data['emotions']['label'] ?? "Unknown",
+                            style: const TextStyle(
+                              fontSize: 18,
+                              fontWeight: FontWeight.w900,
+                              color: Colors.redAccent,
+                            ),
+                          ),
+                          const SizedBox(height: 6),
+
+                          Wrap(
+                            spacing: 6,
+                            children:
+                                (data['emotions']['top5'] as List? ?? [])
+                                    .map<Widget>(
+                                      (e) => Chip(
+                                        label: Text(
+                                          e['name'],
+                                          style: const TextStyle(fontSize: 10),
+                                        ),
+                                        visualDensity: VisualDensity.compact,
+                                      ),
+                                    )
+                                    .toList(),
+                          ),
+                        ],
+                      ),
+                    ),
+
+                  // 6. Lighting Embeddings
+                  // 6. Lighting Embeddings
+                  if (data['lighting'] != null)
+                    _buildSection(
+                      "Lighting Style",
+                      Icons.wb_sunny,
+                      Colors.amber,
+                      Column(
+                        crossAxisAlignment: CrossAxisAlignment.start,
+                        children: [
+                          
+                          Text(
+                            data['lighting']['label'] ?? "Unknown",
+                            style: const TextStyle(
+                              fontSize: 18,
+                              fontWeight: FontWeight.w900,
+                              color: Colors.amber,
+                            ),
+                          ),
+                          const SizedBox(height: 6),
+
+                          Wrap(
+                            spacing: 6,
+                            children:
+                                (data['lighting']['top5'] as List? ?? [])
+                                    .map<Widget>(
+                                      (e) => Chip(
+                                        label: Text(
+                                          e['name'],
+                                          style: const TextStyle(fontSize: 10),
+                                        ),
+                                        visualDensity: VisualDensity.compact,
+                                      ),
+                                    )
+                                    .toList(),
+                          ),
+                        ],
+                      ),
+                    ),
+
                 ],
               ),
             ),
diff --git a/linux/flutter/generated_plugins.cmake b/linux/flutter/generated_plugins.cmake
@@ -8,7 +8,6 @@ list(APPEND FLUTTER_PLUGIN_LIST
 )
 
 list(APPEND FLUTTER_FFI_PLUGIN_LIST
-  onnxruntime
 )
 
 set(PLUGIN_BUNDLED_LIBRARIES)
diff --git a/pubspec.lock b/pubspec.lock
@@ -9,14 +9,6 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "4.0.7"
-  archive:
-    dependency: transitive
-    description:
-      name: archive
-      sha256: "2fde1607386ab523f7a36bb3e7edb43bd58e6edaf2ffb29d8a6d578b297fdbbd"
-      url: "https://pub.dev"
-    source: hosted
-    version: "4.0.7"
   async:
     dependency: transitive
     description:
@@ -292,26 +284,26 @@ packages:
     dependency: transitive
     description:
       name: leak_tracker
-      sha256: "6bb818ecbdffe216e81182c2f0714a2e62b593f4a4f13098713ff1685dfb6ab0"
+      sha256: "33e2e26bdd85a0112ec15400c8cbffea70d0f9c3407491f672a2fad47915e2de"
       url: "https://pub.dev"
     source: hosted
-    version: "10.0.9"
+    version: "11.0.2"
   leak_tracker_flutter_testing:
     dependency: transitive
     description:
       name: leak_tracker_flutter_testing
-      sha256: f8b613e7e6a13ec79cfdc0e97638fddb3ab848452eff057653abd3edba760573
+      sha256: "1dbc140bb5a23c75ea9c4811222756104fbcd1a27173f0c34ca01e16bea473c1"
       url: "https://pub.dev"
     source: hosted
-    version: "3.0.9"
+    version: "3.0.10"
   leak_tracker_testing:
     dependency: transitive
     description:
       name: leak_tracker_testing
-      sha256: "6ba465d5d76e67ddf503e1161d1f4a6bc42306f9d66ca1e8f079a47290fb06d3"
+      sha256: "8d5a2d49f4a66b49744b23b018848400d23e54caf9463f4eb20df3eb8acb2eb1"
       url: "https://pub.dev"
     source: hosted
-    version: "3.0.1"
+    version: "3.0.2"
   lints:
     dependency: transitive
     description:
@@ -340,10 +332,10 @@ packages:
     dependency: transitive
     description:
       name: meta
-      sha256: e3641ec5d63ebf0d9b41bd43201a66e3fc79a65db5f61fc181f04cd27aab950c
+      sha256: "23f08335362185a5ea2ad3a4e597f1375e78bce8a040df5c600c8d3552ef2394"
       url: "https://pub.dev"
     source: hosted
-    version: "1.16.0"
+    version: "1.17.0"
   mime:
     dependency: transitive
     description:
@@ -360,14 +352,6 @@ packages:
       url: "https://pub.dev"
     source: hosted
     version: "1.0.0"
-  onnxruntime:
-    dependency: "direct main"
-    description:
-      name: onnxruntime
-      sha256: e77ec05acafc135cc5fe7bcdf11b101b39f06513c9d5e9fa02cb1929f6bac72a
-      url: "https://pub.dev"
-    source: hosted
-    version: "1.4.1"
   path:
     dependency: "direct main"
     description:
@@ -625,10 +609,10 @@ packages:
     dependency: transitive
     description:
       name: test_api
-      sha256: fb31f383e2ee25fbbfe06b40fe21e1e458d14080e3c67e7ba0acfde4df4e0bbd
+      sha256: ab2726c1a94d3176a45960b6234466ec367179b87dd74f1611adb1f3b5fb9d55
       url: "https://pub.dev"
     source: hosted
-    version: "0.7.4"
+    version: "0.7.7"
   typed_data:
     dependency: transitive
     description:
@@ -649,10 +633,10 @@ packages:
     dependency: transitive
     description:
       name: vector_math
-      sha256: "80b3257d1492ce4d091729e3a67a60407d227c27241d6927be0130c98e741803"
+      sha256: d530bd74fea330e6e364cda7a85019c434070188383e1cd8d9777ee586914c5b
       url: "https://pub.dev"
     source: hosted
-    version: "2.1.4"
+    version: "2.2.0"
   vm_service:
     dependency: transitive
     description:
@@ -686,5 +670,5 @@ packages:
     source: hosted
     version: "6.5.0"
 sdks:
-  dart: ">=3.7.2 <4.0.0"
+  dart: ">=3.8.0-0 <4.0.0"
   flutter: ">=3.29.0"
diff --git a/pubspec.yaml b/pubspec.yaml
@@ -23,12 +23,6 @@ dependencies:
   image: ^4.5.4
   flutter_onnxruntime: ^1.6.0
 
-  onnxruntime: ^1.2.0
-  image: ^4.5.4
-
-  onnxruntime: ^1.2.0
-  image: ^4.5.4
-
 dev_dependencies:
   flutter_test:
     sdk: flutter
@@ -39,17 +33,14 @@ flutter:
   uses-material-design: true
 
   assets:
-      - assets/clip_vitb32_image.onnx
-      - assets/emotion_centroids.json
-      - assets/lighting_centroids.json
-
-  assets:
     - assets/fonts/
     - assets/dinov2.onnx
     - assets/dinov2.onnx.data
     - assets/texture_centroids.json
     - assets/clip_image_encoder.onnx
     - assets/style_embeddings.json
+    - assets/emotion_centroids.json
+    - assets/lighting_centroids.json
     
   fonts:
     - family: GeneralSans
@@ -63,4 +54,4 @@ flutter:
         - asset: assets/fonts/GeneralSans-Medium.otf
           weight: 500
         - asset: assets/fonts/GeneralSans-Semibold.otf
-          weight: 600
+          weight: 600
+\ No newline at end of file
diff --git a/windows/flutter/generated_plugins.cmake b/windows/flutter/generated_plugins.cmake
@@ -8,7 +8,6 @@ list(APPEND FLUTTER_PLUGIN_LIST
 )
 
 list(APPEND FLUTTER_FFI_PLUGIN_LIST
-  onnxruntime
 )
 
 set(PLUGIN_BUNDLED_LIBRARIES)