Eglantine Fonrose

I am a fifth-year engineering student in Audiovisual and Multimedia (computer science applied to video, image and multimedia) at INSA Hauts-de-France, and I am looking for an internship of at least 5 months starting in February 2027.

During my internship at INA, I compared transcription models (including Whisper) and evaluated their integration into FFMPEG for real-time subtitling, analyzing the performance/quality trade-offs. I also designed a video similarity detection method using SSIM and an audio similarity method using MFCC.

In my personal project Groovy Morning FM – RLAC, I use AI to segment radio streams in real time: transcription (Whisper/Kyutai), detection of the start of radio segments by LLM, and training of audio and text models (training of non-pretrained Random Forest models, fine-tuning of Wav2Vec2, AST and CamemBERT, with experiment tracking via Weights & Biases).

INA (French National Audiovisual Institute) — Audiovisual and IT Internship (September 2025, January 2026)