Sync Preprod with main

README.md

@@ -1,7 +1,27 @@
 # Audemy
 
-Audemy is an educational platform dedicated to empowering blind and visually impaired learners by providing accessible and interactive resources.
+Audemy is a 501(c)(3) nonprofit organization reimagining education through the lens of accessibility, inclusion, and intelligent design. We build open-source tools, creative media, and assistive technologies that empower blind and visiually impaired learners.
 
-Supported by a global team of 30 volunteers, Audemy has already impacted over 1,000 blind students across 7 schools worldwide, promoting independence, creativity, and a love for learning. 
+Our mission is to ensure that every child, regardless of ability or background, has access to joyful, high-quality education.
 
-As part of its mission, Audemy continues to innovate in educational assistive technology, proving that tech has the power to create equal opportunities for everyone.
+OUR REACH:
+🎧 10,000+ blind and low-vision students served
+
+🌍 15+ countries, including the U.S., India, Brazil, Nigeria, and the Philippines
+
+🧑‍🤝‍🧑 60+ volunteers powering research, design, and deployment
+
+💵 $200,000 raised from Intel, Taco Bell Foundation, NCWIT, and local scholarships
+
+
+Audemy’s educational games are:
+
+🎙️ Audio-based: Designed for screen-free learning
+
+🧠 AI-enhanced: Personalized with generative voice AI
+
+🎶 Emotionally responsive: Games that laugh, cheer, and encourage
+
+🧩 Curriculum-aligned: Covers foundational math, science, and language
+
+ Visit audemy.org to learn more, or reach out at crystal [at] audemy [dot] org

src/Utilities/openvinoSpeechRec

@@ -0,0 +1,112 @@
+"""
+Copyright (c) 2018-2024 Intel Corporation
+
+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.
+"""
+
+from ..representation import (
+    CharacterRecognitionAnnotation,
+    CharacterRecognitionPrediction,
+)
+from .metric import PerImageEvaluationMetric
+from .distance import editdistance_eval
+
+
+class SpeechRecognitionWER(PerImageEvaluationMetric):
+    __provider__ = 'wer'
+    annotation_types = (CharacterRecognitionAnnotation,)
+    prediction_types = (CharacterRecognitionPrediction,)
+
+    def configure(self):
+        self.words = 0
+        self.score = 0
+
+    def update(self, annotation, prediction):
+        cur_score = editdistance_eval(annotation.label.split(), prediction.label.split())
+        cur_words = len(annotation.label.split())
+        self.score += cur_score
+        self.words += cur_words
+        return cur_score / cur_words if cur_words != 0 else 0
+
+    def evaluate(self, annotations, predictions):
+        return self.score / self.words if self.words != 0 else 0
+
+    def reset(self):
+        self.words, self.score = 0, 0
+
+    @classmethod
+    def get_common_meta(cls):
+        meta = super().get_common_meta()
+        meta['target'] = 'higher-worse'
+        return meta
+
+
+class SpeechRecognitionCER(PerImageEvaluationMetric):
+    __provider__ = 'cer'
+    annotation_types = (CharacterRecognitionAnnotation,)
+    prediction_types = (CharacterRecognitionPrediction,)
+
+    def configure(self):
+        self.length = 0
+        self.score = 0
+
+    def update(self, annotation, prediction):
+        cur_score = editdistance_eval(annotation.label, prediction.label)
+        cur_length = len(annotation.label)
+        self.score += cur_score
+        self.length += cur_length
+        return cur_score / cur_length if cur_length != 0 else 0
+
+    def evaluate(self, annotations, predictions):
+        return self.score / self.length if self.length != 0 else 0
+
+    def reset(self):
+        self.length, self.score = 0, 0
+
+    @classmethod
+    def get_common_meta(cls):
+        meta = super().get_common_meta()
+        meta['target'] = 'higher-worse'
+        return meta
+
+
+class SpeechRecognitionSER(PerImageEvaluationMetric):
+    __provider__ = 'ser'
+
+    annotation_types = (CharacterRecognitionAnnotation,)
+    prediction_types = (CharacterRecognitionPrediction,)
+
+    def configure(self):
+        self.length = 0
+        self.score = 0
+
+    def update(self, annotation, prediction):
+        # remove extra whitespaces
+        gt_label = ' '.join(annotation.label.split())
+        pred_label = ' '.join(prediction.label.split())
+        ser = int(gt_label != pred_label)
+        self.score += ser
+        self.length += 1
+        return ser
+
+    def evaluate(self, annotations, predictions):
+        return self.score / self.length if self.length != 0 else 0
+
+    def reset(self):
+        self.length, self.score = 0, 0
+
+    @classmethod
+    def get_common_meta(cls):
+        meta = super().get_common_meta()
+        meta['target'] = 'higher-worse'
+        return meta