跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 28 天前AI 評分35

ChronoReplay:單一答案介面不足以應對串流空間智能

ChronoReplay: One Answer Interface Is Not Enough for Streaming Spatial Intelligence

AI 導讀

ChronoReplay 是提升串流視覺語言模型空間推理能力的訓練框架,旨在解決模型受單一答案格式限制、難以應對不同空間推理任務的問題。它將不同類型的空間監督資料整理成按時間排序的互動片段,保留各任務的答案格式並平衡不同監督類別。

正文

ChronoReplay is a training framework designed to improve the spatial reasoning capabilities of streaming vision-language models (VLMs).

Existing spatial VLMs often specialize in specific answer formats, such as categorical spatial relations, metric distances, or motion trajectories. This specialization, which we call interface lock-in, limits generalization across different spatial reasoning tasks.

To address this challenge, ChronoReplay organizes heterogeneous spatial supervision into chronologically ordered episodes, preserves task-specific answer formats, and balances learning across supervision families. This enables a single model to learn complementary spatial capabilities while maintaining temporal consistency.

Key Highlights:

Unified Spatial Learning: Supports categorical, metric, and trajectory-based spatial reasoning within a single model.

Chronological Replay: Preserves temporal dependencies by organizing training examples into ordered interaction episodes.

Family-Balanced Training: Improves cross-task generalization by balancing heterogeneous supervision signals.

Strong Benchmark Performance: Achieves consistent improvements across OVO-S-Bench, OST-Bench, and UCS-Bench, including a 13.41-point improvement on OST-Bench.

Efficient Streaming Inference: Adaptive frame selection further improves performance while reducing visual processing costs.

ChronoReplay demonstrates that effective streaming spatial intelligence requires preserving complementary spatial information across tasks, learning from temporally structured experiences, and selecting relevant visual evidence at inference time.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co