2025 GAITE Extras
🎙️ Synthetic Speech Detector
Spot AI-generated speech from spectrograms — the statement literally tells you the solution.
📜The task, in plain English
Classify speech clips as real or AI-generated. The data comes as Mel spectrograms already prepared as tensors — the audio has been turned into images for you.
| Input | Mel spectrogram tensors |
|---|---|
| Output | Real vs synthetic (binary) |
| Really tests | Whether you read the statement: it says treat it as image classification, ResNet18 is enough, ~1 epoch |
🔧The baseline you're given
A starter classification pipeline on the spectrogram tensors. The GAITE statement itself supplies the plan — this task is the purest example of "the hints are the intended solution."
🚀Baseline vs. solution
⚠️ The baseline (what you are given)
- Under-trained starter model
- Spectrogram channels/normalization handled naively
✅ The winning approach
- Do what the statement says: ResNet18, adapt input channels, ~1 epoch
- Normalize spectrograms; check class balance; hold out a validation split
- Time saved here goes to harder tasks — recognizing an "easy by design" task IS the skill
🧒Explain it like I'm brand new
Your Week-3 checkpoint task. It combines everything gently: spectrograms (weird data as images), transfer learning (ResNet18 recipe), a binary metric, and a submission file. When this takes you under two hours end-to-end, your PyTorch literacy is contest-ready.
💬The Gemma 4 playthrough (2000-token limit)
Chat 1 · Full recipe in 3 asks
YOU
Ask 1: dataloaders from tensors X (N,1,128,256), y (N,) — ONLY the Dataset/DataLoader code. Ask 2 (fresh chat): resnet18 adapted to 1-channel input, 2 classes — ONLY the model surgery. Ask 3 (fresh chat): training loop, 1 epoch, AdamW, print val accuracy — ONLY the loop.
Three fresh chats, three small pastes, assemble in your notebook. This is the chunked-prompting rhythm to make automatic before contest day.
🎯Takeaways & what Day 1 might do with this
- Pattern family: spectrograms as images + transfer learning (direct warm-up for 2026 Night Watch).
- Points-per-hour triage: bank easy tasks fast, spend the surplus where it pays.