paper-with-me

Papers

Rethinking Position Embedding as a Context Controller for Multi-Reference and Multi-Shot Video Generation

2026-04-04 · Binyuan Huang, Yuning Lu, Weinan Jia, Hualiang Wang, Mu Liu, Daiqing Yang arxiv

Recent proprietary models such as Sora2 demonstrate promising progress in generating multi-shot videos conditioned on multiple reference characters. However, academic research on this problem remains limited. We study this task and identify a core challenge: when reference images exhibit highly similar appearances, the model often suffers from reference confusion, where semantically similar tokens degrade the model's ability to retrieve the correct context. To address this, we introduce PoCo (Position Embedding as a Context Controller), which incorporates position encoding as additional context control beyond semantic retrieval. By employing side information of tokens, PoCo enables precise token-level matching while preserving implicit semantic consistency modeling. Building on PoCo, we develop a multi-reference and multi-shot video generation model capable of reliably controlling characters with extremely similar visual traits. Extensive experiments demonstrate that PoCo improves cross-shot consistency and reference fidelity compared with various baselines.

📄 PDF Abstract BibTeX arXiv:2604.03738

Code (0)

등록된 구현이 없습니다.

Tasks

Semantic RetrievalVideo Generation

Similar Papers 제목 키워드 기반

Rethinking Addressing in Language Models via Contexualized Equivariant Positional Encoding

2025-01-01 · Jiajun Zhu, Peihao Wang, Ruisi Cai, Jason D. Lee 외

Transformers rely on both content-based and position-based addressing mechanisms to make predictions, but existing positional encoding techniques often diminish the effectiveness of position-based addressing. Many curren…

Arithmetic ReasoningLanguage ModelingLanguage Modellingparameter-efficient fine-tuning+1

Rethinking Homogeneity of Vision and Text Tokens in Large Vision-and-Language Models

2025-02-04 · Chia-Wen Kuo, Sijie Zhu, Fan Chen, Xiaohui Shen 외

Large vision-and-language models (LVLMs) typically treat visual and textual embeddings as homogeneous inputs to a large language model (LLM). However, these inputs are inherently different: visual inputs are multi-dimens…

Language ModelingLanguage ModellingLarge Language Model

Rethinking Positional Encoding in Language Pre-training

2020-06-28 · ICLR 2021 1 · Guolin Ke, Di He, Tie-Yan Liu

In this work, we investigate the positional encoding methods used in language pre-training (e.g., BERT) and identify several problems in the existing formulations. First, we show that in the absolute positional encoding,…

Natural Language UnderstandingSentenceWord Embeddings

Rethinking Positional Encoding

2021-07-06 · Jianqiao Zheng, Sameera Ramasinghe, Simon Lucey

It is well noted that coordinate based MLPs benefit -- in terms of preserving high-frequency information -- through the encoding of coordinate positions as an array of Fourier features. Hitherto, the rationale for the ef…

From Symbolic to Natural-Language Relations: Rethinking Knowledge Graph Construction in the Era of Large Language Models

2026-01-14 · Kanyao Han, Yushang Lai arxiv

Knowledge graphs (KGs) have commonly been constructed using predefined symbolic relation schemas, typically implemented as categorical relation labels. This design has notable shortcomings: real-world relations are often…

Knowledge Graphs