MELD

by SenticNet group / NUS

Multimodal emotion corpus from Friends TV show — conversational emotion recognition.

TL;DR

Multimodal emotion corpus from Friends TV show — conversational emotion recognition.

Best for conversational emotion recognition; multimodal (audio + video + text). Pricing: research-only.

Category
Open source
License
—
Stars
—
Last push
—
Pricing
research-only
Platforms
GitHub

What it is

MELD extends EmotionLines with audio + video from Friends TV show. 13k utterances across 7 emotion classes in conversation. Audio/video copyright-bound.

Best for: Conversational emotion recognition; multimodal (audio + video + text).
Watch out for: GPL-3.0 metadata · audio-visual content from Friends TV show is COPYRIGHTED · research-only fair-use claim. Cite: Poria et al., ACL 2019.

Install / use

git clone https://github.com/declare-lab/MELD

Features

Speaker diarizationYes
Word-level timestampsNo
Streaming / real-timeNo
Languages supported1
HIPAA eligibleNo

MELD vs Whipscribe

FeatureMELDWhipscribe
CategoryOpen sourceTranscription APIs
Pricingresearch-onlyfree beta
Speaker diarizationYesYes
Word timestampsNoYes
StreamingNoNo
Languages199
PlatformsGitHubWeb, API, MCP

Alternatives to MELD

OpenAI Whisper
OpenAI

The reference open-source multilingual ASR model from OpenAI.

OSS · MIT ★ 98.1k
whisper.cpp
Georgi Gerganov

C/C++ port of Whisper — runs on anything, from a Raspberry Pi to Apple Silicon.

OSS · MIT ★ 48.8k
faster-whisper
SYSTRAN

4× faster than reference Whisper using CTranslate2 — production sweet spot.

OSS · MIT ★ 22.3k

Whipscribe is a managed faster-whisper + whisperX service. If you want transcripts without running infrastructure, paste a URL or drop a file in the form below — you'll have a transcript in seconds.