@adamrcmmm33630

Echo is an automated video processing, multi-persona captioning, and LLM-Judge evaluation pipeline. The system ingests short-form video files and utilizes OpenCV to extract sparse, optimized keyframes across the clip's duration. These frames are dispatched to a multimodal vision model (Qwen-3.7-Plus) to generate a dense, high-fidelity visual scene description. This visual context is then orchestrated through a deep-reasoning language model (DeepSeek-V4-Pro) across four distinct, concurrent style tracks: Formal (objective summary), Sarcastic (cynical irony), Humorous-Tech (software engineering metaphors), and Humorous-Non-Tech (relatable stand-up observational comedy). To bypass conversational filler from the reasoning model, custom regex parsing filters isolate the raw output within strict XML container tags. Finally, an independent automated LLM-Judge scores each variation on a 1–5 scale for Accuracy and Tone Alignment. The entire framework is exposed through a reactive Streamlit dashboard providing upload handling, live pipeline status execution logs, and side-by-side results tabs.

To participate in the hackathon, simply click Sign up with AMD. If you are not already a member of the AMD AI Developer Program (ADP), you will need to create an account. ADP members can access AI tutorials, experts, and community support. In addition, new members can claim a free month of DeepLearning.ai pro, $100 in AMD GPU cloud credits, and $50 in Fireworks AI API credits. All hackathon participants will receive on hackathon launch additional compute and API credits for their projects. Ready to build what's next on AMD? Sign up with AMD below to secure your spot.