WORLDWIDE_SanGuai / Research bulletin

[ worldwide.arakne.de ]

A big world. A little reading.

A three-day research bulletin by SanGuai. Drop in, read, move on.

EST. 202672H72 HOURS ON THE BOARD
Artificial intelligence / 2611.14601

VASTRA: Efficient Agentic Top-K Interactive Video Analysis

Authors: Tushar Tyagi, Aditya Raghavan, Haoran Qiu, Francisco RomeroPaper published: 2026-11-13Leaves in: 16 hours

THE QUICK READ

The authors present VASTRA, an agentic video retrieval system that lets query complexity, not an open-ended reasoning loop, drive execution: easy queries are ranked from pre-ingested summaries, while harder ones trigger targeted verification only where relevance is uncertain. The abstract reports results on 40 benchmark queries, including a fast preset that is 12.2x quicker and 10.4x cheaper than the best ReAct agent at a 5.0-point mAP reduction.

KEY POINTS

  1. Control flow is kept external to the model rather than latent in a growing chain of thought, which the authors say yields deterministic control flow by construction.
  2. VASTRA-Fast runs in 4.2 s at $0.0044 versus 51.2 s and $0.0458 for the best ReAct agent across 40 benchmark queries.
  3. VASTRA-Max reaches 0.653 mAP versus 0.649 for the ReAct agent, which the authors describe as comparable accuracy with zero run-to-run variance but higher spend.

READING LIMITS

Based only on the title and abstract; the method, benchmark composition, cost accounting and peer-review status were not examined, and all figures are the authors' own claims without independent verification.

Source abstract [EN]

Vision-language models (VLMs) enable natural language video retrieval, bypassing traditional hand-engineered pipelines. Integrating agentic reasoning frameworks (e.g., ReAct) further advances this paradigm by allowing models to dynamically inspect video content rather than relying on a single visual pass. Applying this pattern naively to video retrieval inherits problems that video systems cannot afford: results vary from run to run under identical conditions, and cost and latency scale with every reasoning step the agent takes rather than with the work the query actually requires. We present VASTRA, an agentic video retrieval system whose execution is dictated by query complexity rather than an open-ended reasoning loop: cheap, reliably answerable queries are ranked directly from pre-ingested summaries, while harder queries trigger targeted verification only where relevance is genuinely uncertain. This keeps control flow external to the model rather than latent in a growing chain of thought, establishing deterministic control flow by construction and end-to-end rank reproducibility under greedy serving. Across 40 benchmark queries, VASTRA exposes explicit presets: VASTRA-Fast runs 12.2x faster (4.2 s vs. 51.2 s) and 10.4x cheaper ($0.0044 vs. $0.0458) than best performing ReAct agent at a 5.0-point mAP reduction, while VASTRA-Max matches peak ReAct agent accuracy (0.653 vs. 0.649 mAP) at comparable latency, trading higher spend ($0.1022 vs. $0.0458) for zero run-to-run variance.

AI summaries use public titles and abstracts only. Research claims belong to the authors; consult the original paper.