Curated insights and tools for curious minds.

📅 Date: 10/03/2025

⏱️Read time: 3 minutes

1. 🎯 Highlight of the Week

• ✨ Resource: How I Use LLMs
• 💡 Why it matters: LLMs are everywhere these days, helping with all sorts of tasks. With so many options, it’s worth figuring out which one works best for what.

2. 📚 Key Articles & Insights

This week, I’ve been exploring how to use sound signal processing and generative AI in my research. I looked into audio feature extraction, processing (seems like librosa in Python is a go-to), enhancing with gen AI, and classification. Found some solid review article along the way—listing it below!

• 📰 Enhanced Feature Extraction Approaches for Detection of Sound Events – The paper explores digital signal processing techniques to analyze and extract audio features for predicting sound events using supervised machine learning. It evaluates five classification approaches using different subsets of audio features.

N. Surampudi, M. Srirangan and J. Christopher, "Enhanced Feature Extraction Approaches for Detection of Sound Events," 2019 IEEE 9th International Conference on Advanced Computing (IACC), Tiruchirappalli, India, 2019, pp. 223-229, doi: 10.1109/IACC48062.2019.8971574.

3. ⚙️ Tools & Resources

• 🔧 Structuring Your PhD Thesis: The Power of Storytelling:

During my meetings with my PhD supervisor, he often emphasizes that storytelling and clarity of expression are just as crucial as the technical content itself. If you're about to start writing your PhD thesis, like me, and need help with structuring, I found Emmanuel’s diagram both insightful and widely applicable. While thesis structures may vary across disciplines, dedicating time to storytelling before diving into the writing process provides a strong foundation and a clear direction.

https://www.linkedin.com/posts/emmanueltsekleves_phd-phdlife-thesis-activity-7303770149269000192-n469?utm_source=share&utm_medium=member_desktop&rcm=ACoAAAl9_msBEjRrUBT8yxJzHEk0WrhdSllWta0

• 🛠️ How I Use LLMs: If you're interested in exploring the latest capabilities of popular LLMs—such as multimodal inputs, internet search, and Python-integrated responses—Andrej has put together a clear and accessible tutorial.

  Freely available at https://www.youtube.com/watch?v=EWvNQjAaOHw