CASTLE2026 Team WDL Technical Report
This is a practical engineering solution for a specific benchmark challenge, incremental in nature.
The team tackled long-form egocentric video QA by building an evidence-aware multimodal reasoning pipeline based on Qwen, achieving first place in the CASTLE Challenge @ EgoVis 2026 with a final score of 0.58.
The CASTLE Challenge @ EgoVis 2026 evaluates long-form egocentric video question answering over 600+ hours of multi-perspective recordings. Each four-choice question requires evidence from videos, transcripts, auxiliary photos, people, days, rooms, and temporal context. We propose an evidence-aware multimodal reasoning pipeline based on Qwen. Our system parses question hints, retrieves ASR chunks, attaches auxiliary images, samples candidate video frames, and routes questions into static visual, speech/text, temporal, and mixed types with specialized prompts. Multiple inference passes are aggregated by confidence-weighted voting and converted into the official Codabench format. In ablation, LoRA improves the score from 0.21 to 0.50, and more sampled frames further raise it to 0.58. Our final system ranks first in the CASTLE Challenge @ EgoVis 2026.