Výsledky vyhledávání - "Song WenFeng"

Report

Raw Text is All you Need: Knowledge-intensive Multi-turn Instruction Tuning for Large Language Model

Autor: Hou, Xia, Li, Qifeng, Yang, Jian, Li, Tongliang, Chai, Linzheng, Wu, Xianjie, Ji, Hangyuan, Li, Zhoujun, Nie, Jixuan, Dun, Jingbo, Song, Wenfeng

Instruction tuning as an effective technique aligns the outputs of large language models (LLMs) with human preference. But how to generate the seasonal multi-turn dialogues from raw documents for instruction tuning still requires further exploration.

Externí odkaz: http://arxiv.org/abs/2407.03040

Zobrazit plný text záznamu

Report

Rethinking Object Saliency Ranking: A Novel Whole-flow Processing Paradigm

Autor: Song, Mengke, Li, Linfeng, Wu, Dunquan, Song, Wenfeng, Chen, Chenglizhao

Existing salient object detection methods are capable of predicting binary maps that highlight visually salient regions. However, these methods are limited in their ability to differentiate the relative importance of multiple objects and the relation

Externí odkaz: http://arxiv.org/abs/2312.03226

Zobrazit plný text záznamu

Report

A Comprehensive Survey on Video Saliency Detection with Auditory Information: the Audio-visual Consistency Perceptual is the Key!

Autor: Chen, Chenglizhao, Song, Mengke, Song, Wenfeng, Guo, Li, Jian, Muwei

Video saliency detection (VSD) aims at fast locating the most attractive objects/things/patterns in a given video clip. Existing VSD-related works have mainly relied on the visual system but paid less attention to the audio aspect, while, actually, o

Externí odkaz: http://arxiv.org/abs/2206.13390

Zobrazit plný text záznamu