Upload 5 files

.gitignore

@@ -0,0 +1 @@
+venv/

README.md

@@ -1,19 +1,2 @@
----
-title: Echomatch
-emoji: 🚀
-colorFrom: red
-colorTo: red
-sdk: docker
-app_port: 8501
-tags:
-- streamlit
-pinned: false
-short_description: Streamlit template space
----
-
-# Welcome to Streamlit!
-
-Edit `/src/streamlit_app.py` to customize this app to your heart's desire. :heart:
-
-If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
-forums](https://discuss.streamlit.io).
+# echomatch
+Transformer-Based Voice-to-Text Emotion Matching

main.py

@@ -0,0 +1,90 @@
+import os
+import streamlit as st
+import tempfile
+from pydub import AudioSegment
+from app.audio_processing import transcribe_audio
+from app.emotion_text import get_emotion, emotion_labels
+
+# Configure Streamlit app
+st.set_page_config(
+    page_title="EchoMatch: Emotion from Voice",
+    page_icon="🎧",
+    layout= "centered"
+)
+
+st.title("🎧 EchoMatch: Emotion from Voice")
+st.markdown("Upload an audio file (MP3, WAV, M4A, FLAC, OGG) to get a text transcription and emotional analysis.")
+
+uploaded_file = st.file_uploader(
+    "Upload an audio file (max 15MB recommended)",
+    type=["wav", "mp3", "m4a", "flac", "ogg"]
+)
+
+if uploaded_file is None:
+    st.info("Please upload an audio file to get started.")
+else:
+    if uploaded_file.size > 15 * 1024 * 1024:
+        st.warning("File size exceeds the recommended 15MB limit. Processing may be slow or fail for larger files.")
+    
+    st.audio(uploaded_file, format=uploaded_file.type)
+
+    temp_audio_path = None
+
+    try:
+        suffix = os.path.splitext(uploaded_file.name)[1]
+        with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp_file:
+            tmp_file.write(uploaded_file.read())
+            temp_audio_path = tmp_file.name
+
+        if not temp_audio_path.lower().endswith(".wav"):
+            st.info("Converting audio to WAV format for reliable processing...")
+            try:
+                audio = AudioSegment.from_file(temp_audio_path)
+                wav_audio_path = temp_audio_path.replace(suffix, ".wav")
+                audio.export(wav_audio_path, format="wav")
+                os.remove(temp_audio_path)
+                temp_audio_path = wav_audio_path
+            except Exception as e:
+                st.error(f"Could not convert audio to WAV. Please ensure ffmpeg is installed and correctly configured in your system's PATH. Error: {e}")
+                st.stop()
+
+        st.subheader("📜 Transcription")
+        with st.spinner("Transcribing audio... This might take a while for longer files."):
+            transcribed_text = transcribe_audio(temp_audio_path)
+
+        if transcribed_text:
+            st.success("Transcription complete!")
+            st.markdown(f"**Transcribed Text:**\n\n```\n{transcribed_text}\n```")
+        else:
+            st.warning("Could not generate transcription for this audio file.")
+
+        st.subheader("😄 Detected Emotions")
+        if transcribed_text:
+            with st.spinner("Analyzing emotions..."):
+                emotion_scores = get_emotion(transcribed_text)
+            
+            if emotion_scores:
+                st.markdown("Here are the detected emotion scores:")
+                
+                cols = st.columns(len(emotion_labels))
+                for i, label in enumerate(emotion_labels):
+                    with cols[i]:
+                        score_percent = f"{emotion_scores.get(label, 0.0) * 100:.2f}%"
+                        st.metric(label.capitalize(), score_percent)
+                
+                if st.checkbox("Show raw emotion scores (JSON)"):
+                    st.json(emotion_scores)
+            else:
+                st.info("Could not detect emotions from the transcription (it might be empty or too short).")
+        else:
+            st.info("Emotion analysis skipped as transcription was not available.")
+
+    except Exception as e:
+        st.error(f"An unexpected error occurred during processing: {e}")
+        st.exception(e)
+    finally:
+        if temp_audio_path and os.path.exists(temp_audio_path):
+            os.remove(temp_audio_path)
+
+st.markdown("---")
+st.markdown("EchoMatch: Emotion from Voice. Powered by Whisper ASR and Hugging Face Transformers.")

packages.txt

@@ -0,0 +1,2 @@
+ffmpeg
+libportaudio2

requirements.txt

@@ -1,3 +1,8 @@
-altair
-pandas
-streamlit
+streamlit
+git+https://github.com/openai/whisper.git
+pydub
+scipy>=1.7.3
+torch
+torchaudio
+transformers
+ffmpeg-python