Learning AI from scratch– category –

Recap and Today's Theme Hello! In the previous episode, we reviewed and conducted a knowledge check for Chapter 7, reflecting on and deepening our understanding of the concepts we covered. Today, we begin Chapter 8 and explore the basics...

Recap: Diffusion Models In the previous episode, we discussed Diffusion Models, explaining their mechanisms, applications, and differences from other generative models like GANs and VAEs. Diffusion models are gaining attention for their ...

Recap: Challenges and Limitations of Generative Models In the previous episode, we explored the challenges and limitations of generative models, including quality issues, computational costs, and ethical concerns. While generative models...

Recap: Applications of Generative Models In the previous episode, we explored various applications of generative models such as image generation, text generation, and speech synthesis. These technologies have broad applications, ranging ...

Recap: Model Safety and Filtering In the previous episode, we explored filtering techniques to ensure the safety of AI models. Methods such as the blacklist approach, NLP filtering, and Human-in-the-Loop (HITL) were discussed to prevent ...

Recap: Prompt Tuning In the previous episode, we learned about prompt tuning, a technique to optimize prompts for eliciting desired outputs from pre-trained models. Designing prompts effectively is crucial for enhancing model accuracy an...

Recap: Applications of Self-Supervised Learning In the previous episode, we discussed the applications of self-supervised learning, a method that learns features from unlabeled data. This approach has proven valuable in fields such as na...

Recap: Large-Scale Pre-Trained Models In the previous episode, we discussed large-scale pre-trained models, which are models trained on massive datasets that exhibit high performance when their general features are applied to specific ta...

Recap: Evaluation Metrics for Speech Generation In the previous episode, we explained how to evaluate the quality of speech generation using metrics like PESQ and STOI for objective evaluation and MOS for subjective evaluation. These met...

Recap: Tacotron In the previous episode, we explained Tacotron, a model that converts text into speech, widely used in applications such as voice assistants and narration generation. Especially with Tacotron 2, the quality of the generat...

Recap: WaveNet In the previous episode, we explained WaveNet, a neural network-based model that directly generates speech waveforms, offering high-quality speech synthesis. By generating audio at the sample level, WaveNet produces more n...

Recap: Speech Generation Models In the previous episode, we discussed the basics of speech generation technology. Traditional methods such as rule-based, unit selection, and parametric speech synthesis have evolved into modern approaches...

Recap: Evaluation Metrics for Text Generation In the previous episode, we discussed evaluation metrics for text generation, explaining key metrics like Perplexity and the BLEU score. Perplexity measures a model's prediction accuracy, whi...

Recap: BERT and the Masked Language Model In the previous episode, we explored BERT (Bidirectional Encoder Representations from Transformers), a powerful model in natural language processing (NLP), and its training method, the Masked Lan...

Recap: Multi-Head Attention Mechanism In the previous episode, we explained the Multi-Head Attention Mechanism, a core technology within the Transformer model. This mechanism allows the model to understand the context of text from multip...