Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,64 +1,22 @@
|
|
| 1 |
-
# 🎙 AI Conversation Sentiment Analyzer
|
| 2 |
-
|
| 3 |
-
An AI-powered system that transcribes audio, detects speakers, and analyzes sentiment for each utterance.
|
| 4 |
-
|
| 5 |
-
---
|
| 6 |
-
|
| 7 |
-
## 🚀 Features
|
| 8 |
-
|
| 9 |
-
- 🎧 Upload audio or record via microphone
|
| 10 |
-
- 🗣 Automatic speaker detection (multi-speaker support)
|
| 11 |
-
- 💬 Sentiment analysis per utterance
|
| 12 |
-
- ⏱ Timestamped transcription
|
| 13 |
-
- 📁 Export results to:
|
| 14 |
-
- TXT
|
| 15 |
-
- JSON
|
| 16 |
-
- CSV
|
| 17 |
-
- Word (.docx)
|
| 18 |
-
- PDF
|
| 19 |
-
|
| 20 |
-
---
|
| 21 |
-
|
| 22 |
-
## 🧠 How It Works
|
| 23 |
-
|
| 24 |
-
1. Audio is processed using **AssemblyAI** for transcription
|
| 25 |
-
2. Speaker diarization identifies different speakers
|
| 26 |
-
3. Each sentence is analyzed using a **Hugging Face sentiment model**
|
| 27 |
-
4. Results are displayed and can be exported
|
| 28 |
-
|
| 29 |
-
---
|
| 30 |
-
|
| 31 |
-
## ⚙️ Usage
|
| 32 |
-
|
| 33 |
-
1. Upload an audio file or record using the microphone
|
| 34 |
-
2. (Optional) Set number of speakers or leave as `0` for auto
|
| 35 |
-
3. Choose language or keep `auto`
|
| 36 |
-
4. Click **🚀 Analyze**
|
| 37 |
-
5. View and export results
|
| 38 |
-
|
| 39 |
-
---
|
| 40 |
-
|
| 41 |
-
## 📊 Example Output
|
| 42 |
-
|
| 43 |
-
|
| 44 |
---
|
| 45 |
-
title:
|
| 46 |
emoji: 🎙
|
| 47 |
colorFrom: purple
|
| 48 |
colorTo: green
|
| 49 |
sdk: gradio
|
| 50 |
sdk_version: "4.44.0"
|
|
|
|
| 51 |
app_file: app.py
|
| 52 |
pinned: false
|
| 53 |
---
|
| 54 |
|
| 55 |
# AI Conversation Sentiment Analyzer
|
| 56 |
|
| 57 |
-
Upload or record audio
|
| 58 |
|
| 59 |
**Models used:**
|
| 60 |
-
-
|
| 61 |
-
-
|
| 62 |
-
-
|
| 63 |
|
| 64 |
-
**Required secret:** `HF_TOKEN`
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
title: Sentiment Analysis
|
| 3 |
emoji: 🎙
|
| 4 |
colorFrom: purple
|
| 5 |
colorTo: green
|
| 6 |
sdk: gradio
|
| 7 |
sdk_version: "4.44.0"
|
| 8 |
+
python_version: "3.10"
|
| 9 |
app_file: app.py
|
| 10 |
pinned: false
|
| 11 |
---
|
| 12 |
|
| 13 |
# AI Conversation Sentiment Analyzer
|
| 14 |
|
| 15 |
+
Upload or record audio to get speaker-separated transcription with sentiment labels.
|
| 16 |
|
| 17 |
**Models used:**
|
| 18 |
+
- openai/whisper-base — transcription
|
| 19 |
+
- pyannote/speaker-diarization-3.1 — speaker diarization
|
| 20 |
+
- nlptown/bert-base-multilingual-uncased-sentiment — sentiment analysis
|
| 21 |
|
| 22 |
+
**Required secret:** Set `HF_TOKEN` under Space Settings → Repository secrets.
|