#ChenFei
LightNews LightNews
sunlightwang.bsky.social
We wrote a preview for MACS3, which is the most updated version of MACS for epigenetic signal peak calling. The preview article discusses integrative analysis environment, single-cell peak calling strategy, and AI-ready data foundations.
academic.oup.com/gpb/advance-...
The Development of the MACS Lineage: from Peak Calling to Regulatory Genomics in the Single-cell Era
Xi Wang, Chenfei Wang; The Development of the MACS Lineage: from Peak Calling to Regulatory Genomics in the Single-cell Era, Genomics, Proteomics & Bioinfo
academic.oup.com
October 27, 2026 at 4:50 AM
arxiv-cs-ro.bsky.social
Fengkai Liu, Hao Su, Haozhuang Chi, Rui Geng, Congzhi Ren, Xuqing Liu, Chenfei Xu, Yuichi Ohsita, Liyun Zhang
Event-Driven Proactive Robot Assistance through Vision-Language Reasoning
https://arxiv.org/abs/2610.08344
October 8, 2026 at 7:00 AM
csro-bot.bsky.social
Fengkai Liu, Hao Su, Haozhuang Chi, Rui Geng, Congzhi Ren, Xuqing Liu, Chenfei Xu, Yuichi Ohsita, Liyun Zhang: Event-Driven Proactive Robot Assistance through Vision-Language Reasoning https://arxiv.org/abs/2610.08344 https://arxiv.org/pdf/2610.08344 https://arxiv.org/html/2610.08344
October 7, 2026 at 6:45 AM
earie-news.bsky.social
👏 A big thank you to IJIO’s Top Reviewers of 2025!
Luis Aguiar, Vivek Bhattacharya, Jeff Gortmaker, Carl-Christian Groh, Hyo Kang, Yunmi Kong, Marleen Marra, Chenfei Qu, Andrew Rhodes, Luca Sandrini, Yucheng Wang & Cédric Wasser.
October 5, 2026 at 9:43 AM
eesssp-bot.bsky.social
Chenfei Huang, Mingjie Shao, Ya-Feng Liu: Efficient Transmit Waveform Design for MIMO-OFDM DFRC Systems with 1-Bit DACS https://arxiv.org/abs/2609.36745 https://arxiv.org/pdf/2609.36745 https://arxiv.org/html/2609.36745
September 30, 2026 at 6:46 AM
ezc-group.bsky.social
After 2 years in review, very happy to see this paper (doi.org/10.1002/adom... ) finally accepted in Adv. Optical Mater. 🎉Congrats to Chenfei & Changfeng for their contributions to this collab/w the groups of Jacquemin and Han Sen Soo.👏🏻 @st-andrewschem.bsky.social @standrewsosc.bsky.social
September 24, 2026 at 10:59 AM
cshc-bot.bsky.social
Wenjuan Zhong, Chenfei Ma, Kianoush Nazarpour: ViMoWear: Visual Motion-Guided sEMG-IMU Representation Learning for Subject-Independent Thumb Gesture Recognition https://arxiv.org/abs/2609.27595 https://arxiv.org/pdf/2609.27595 https://arxiv.org/html/2609.27595
September 24, 2026 at 6:41 AM
voxdev.bsky.social
🆕 China's nationwide carbon market: Costs, environmental gains, and potential reforms

Today on VoxDev w/ Lawrence H. Goulder (Stanford), Xianling Long (Peking University), Chenfei Qu (Stanford Energy) & Da Zhang (Tsinghua University): https://ow.ly/Rqj450ZBmJs
China's nationwide carbon market: Costs, environmental gains, and potential reforms
China's carbon market is projected to deliver benefits worth more than 20 times its cost – and smarter benchmark and auction design could make it even more effective.
ow.ly
August 19, 2026 at 9:02 AM
paper.bsky.social
[26/30] 251 Likes, 2 Comments, 2 Posts
2607.27205, cs․CV | cs․RO, 29 Jul 2026

🆕TurboVLA: Real-Time Vision-Language-Action Model at 32 Hz on an RTX 4090 with <1 GB VRAM

Hengyi Xie, Chenfei Yao, Xianjin Wu, Xuanyang Xi, Yiping Tang, Di Xu, Yingying Zhu, Dingkang Liang, Xiang Bai, Han Ding
Vision-language-action (VLA) models commonly adopt an LLM-centric $V \to L \to A$ pathway, where visual observations are projected into the representation space of a large language model before being decoded into robot actions. Although effective, this design incurs substantial computation and memory overhead at every policy invocation. In this work, we introduce TurboVLA, a new VLA paradigm that reformulates the conventional $V \to L \to A$ pathway as a direct $V + L \to A$ mapping. Instead of using a large language model as the central interface between perception and action, TurboVLA independently encodes visual observations and language instructions, directly exchanges information between them through lightweight bidirectional vision-language interaction, and predicts continuous action chunks with a compact decoder. This simple design constructs task-conditioned representations directly from visual and linguistic features, significantly reducing the computational and memory costs of VLA inference. On LIBERO, TurboVLA achieves 97.7% average success with only 0.2B parameters, 31.2 ms inference latency, and 0.9 GB inference VRAM on a consumer-grade RTX 4090, matching or outperforming substantially larger VLA policies. These results establish TurboVLA as a simple and effective alternative to the prevailing LLM-centric VLA paradigm, offering a new perspective on how vision, language, and action can be connected for efficient robotic manipulation. Code is available at https://github.com/H-EmbodVis/TurboVLA.
August 15, 2026 at 12:08 AM
akyd.bsky.social
アーティストに描いた絵を見てもらうという（！）イベントにも参加した。かわいがっていた犬の絵を描いた。当たり前のように美術館で絵を観てはいるが、自分が描くことはなかったのでやってみて良かった。稚拙な出来だが描いた絵には自分で描いたということの満足感/特別感があった。チェン・フェイ氏のサイン、犬のイラストを付けてくれたのは嬉しかったな。
#チェンフェイ展 #chenfei
July 12, 2025 at 2:25 PM
akyd.bsky.social
昨日は仕事を早く済ませて、チェン・フェイ展「父と子」（ワタリウム美術館）。
赤ちゃんの排便が素敵にレイアウトされた「ラブレター」、繰り返しの差異が楽しい「生き物のドラマ」にグッときた。
www.watarium.co.jp/jp/exhibitio...
#チェンフェイ展 #chenfei
July 12, 2025 at 2:22 PM
tsurugami.bsky.social
Honkai Star Rail oc, Chenfei shenanigans with Jiaoqiu

#HSR_oc #tsuruocs #oc
October 17, 2024 at 2:42 PM
paper.bsky.social
[19/30] 224 Likes, 7 Comments, 3 Posts
2605.10730, cs․CV, 11 May 2026

🆕Qwen-Image-2.0 Technical Report

Bing Zhao, Chenfei Wu, Deqing Li, Hao Meng, Jiahao Li, Jie Zhang, Jingren Zhou, Junyang Lin, Kaiyuan Gao, Kuan Cao, Kun Yan, Liang Peng, Lihan Jiang, Niantong Li, Ningyuan Tang, Shengming Yi...
We present Qwen-Image-2.0, an omni-capable image generation foundation model that unifies high-fidelity generation and precise image editing within a single framework. Despite recent progress, existing models still struggle with ultra-long text rendering, multilingual typography, high-resolution photorealism, robust instruction following, and efficient deployment, especially in text-rich and compositionally complex scenarios. Qwen-Image-2.0 addresses these challenges by coupling Qwen3-VL as the condition encoder with a Multimodal Diffusion Transformer for joint condition-target modeling, supported by large-scale data curation and a customized multi-stage training pipeline. This enables strong multimodal understanding while preserving flexible generation and editing capabilities. The model supports instructions of up to 1K tokens for generating text-rich content such as slides, posters, infographics, and comics, while significantly improving multilingual text fidelity and typography. It also enhances photorealistic generation with richer details, more realistic textures, and coherent lighting, and follows complex prompts more reliably across diverse styles. Extensive human evaluations show that Qwen-Image-2.0 substantially outperforms previous Qwen-Image models in both generation and editing, marking a step toward more general, reliable, and practical image generation foundation models.
May 18, 2026 at 12:07 AM
paper.bsky.social
[26/30] 251 Likes, 2 Comments, 2 Posts
2508.02324, cs․CV, 04 Aug 2025

🆕Qwen-Image Technical Report

Chenfei Wu, Jiahao Li, Jingren Zhou, Junyang Lin, Kaiyuan Gao, Kun Yan, Sheng-ming Yin, Shuai Bai, Xiao Xu, Yilei Chen, Yuxiang Chen, Zecheng Tang, Zekai Zhang, Zhengyi Wang, An Yang, Bowen Yu, ...
We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address the challenges of complex text rendering, we design a comprehensive data pipeline that includes large-scale data collection, filtering, annotation, synthesis, and balancing. Moreover, we adopt a progressive training strategy that starts with non-text-to-text rendering, evolves from simple to complex textual inputs, and gradually scales up to paragraph-level descriptions. This curriculum learning approach substantially enhances the model's native text rendering capabilities. As a result, Qwen-Image not only performs exceptionally well in alphabetic languages such as English, but also achieves remarkable progress on more challenging logographic languages like Chinese. To enhance image editing consistency, we introduce an improved multi-task training paradigm that incorporates not only traditional text-to-image (T2I) and text-image-to-image (TI2I) tasks but also image-to-image (I2I) reconstruction, effectively aligning the latent representations between Qwen2.5-VL and MMDiT. Furthermore, we separately feed the original image into Qwen2.5-VL and the VAE encoder to obtain semantic and reconstructive representations, respectively. This dual-encoding mechanism enables the editing module to strike a balance between preserving semantic consistency and maintaining visual fidelity. Qwen-Image achieves state-of-the-art performance, demonstrating its strong capabilities in both image generation and editing across multiple benchmarks.
August 8, 2025 at 12:07 AM
paper.bsky.social
[30/30] 121 Likes, 2 Comments, 1 Posts
2505.19147, cs․CL | cs․AI | cs․CV, 25 May 2025

🆕Shifting AI Efficiency From Model-Centric to Data-Centric Compression

Xuyang Liu, Zichen Wen, Shaobo Wang, Junjie Chen, Zhishan Tao, Yubo Wang, Xiangqi Jin, Chang Zou, Yiyu Wang, Chenfei Liao, Xu Zheng, H...
The rapid advancement of large language models (LLMs) and multi-modal LLMs (MLLMs) has historically relied on model-centric scaling through increasing parameter counts from millions to hundreds of billions to drive performance gains. However, as we approach hardware limits on model size, the dominant computational bottleneck has fundamentally shifted to the quadratic cost of self-attention over long token sequences, now driven by ultra-long text contexts, high-resolution images, and extended videos. In this position paper, \textbf{we argue that the focus of research for efficient AI is shifting from model-centric compression to data-centric compression}. We position token compression as the new frontier, which improves AI efficiency via reducing the number of tokens during model training or inference. Through comprehensive analysis, we first examine recent developments in long-context AI across various domains and establish a unified mathematical framework for existing model efficiency strategies, demonstrating why token compression represents a crucial paradigm shift in addressing long-context overhead. Subsequently, we systematically review the research landscape of token compression, analyzing its fundamental benefits and identifying its compelling advantages across diverse scenarios. Furthermore, we provide an in-depth analysis of current challenges in token compression research and outline promising future directions. Ultimately, our work aims to offer a fresh perspective on AI efficiency, synthesize existing research, and catalyze innovative developments to address the challenges that increasing context lengths pose to the AI community's advancement.
May 28, 2025 at 12:05 AM
paper.bsky.social
[9/10] 560 Upvotes, 49 Comments, 2 Posts
https://arxiv.org/abs/2304.10464 cs․CL, 20 Apr 2023

🆕Learning to Program with Natural Language

Yiduo Guo, Yaobo Liang, Chenfei Wu, Wenshan Wu, Dongyan Zhao, Nan Duan
April 23, 2023 at 12:01 AM
sigmod.bsky.social
"FlashANNS: GPU-Driven Asynchronous I/O Pipelining for Eliminating Storage-Compute Bottlenecks in Billion-Scale Similarity Search" by Yang XiaoMo SunZiyu SongBing TianJie SunJie ZhangZeke WangZonghui WangWenzhi ChenFei Wu: dl.acm.org/doi/abs/10.1...
dl.acm.org
April 7, 2026 at 6:10 PM
sigmod.bsky.social
"P2FedRec: Towards Privacy-Preserving and Personalized Federated Recommendation via Relationship Awareness" by Chenfei HuZihao XuTong WuYou LiChuan ZhangLiehuang Zhu: dl.acm.org/doi/abs/10.1...
dl.acm.org
December 6, 2025 at 4:48 AM
repec-nep-env.bsky.social
Did China’s Pilot Emissions Trading Scheme Reduce CO2 Emissions? Evidence from a Meta-Analysis: Yanxia Yu; Chenfei Qu; Tom Coupé; Mathilda Featherston-Lardeux; Andreas Loeschel; Arne R. Weiss; Da Zhang
NEP/RePEc link
to paper
d.repec.org
May 3, 2025 at 6:45 PM
repec-nep-env.bsky.social
Rate-Based Emissions Trading with Overlapping Policies : Insights from Theory and an Application to China: Carolyn Fischer; Qu, Chenfei; Goulder, Lawrence H.
NEP/RePEc link
to paper
d.repec.org
March 26, 2025 at 5:45 AM
repec-nep-env.bsky.social
Rate-Based Emissions Trading with Overlapping Policies: Insights from Theory and an Application to China: Carolyn Fischer; Chenfei Qu; Lawrence H. Goulder
NEP/RePEc link
to paper
d.repec.org
February 4, 2025 at 11:45 PM
babygoldie.bsky.social
[4/4] of the company, and the company will handle industrial and commercial change registration. As of the announcement date, Qin Chenfei did not hold any shares in the company.
June 12, 2026 at 9:55 AM
babygoldie.bsky.social
[2/4] of Qin Chenfei as a non-independent director of the 11th Board of Directors. The ninth meeting of the 11th Board of Directors was held on the same day, and all directors unanimously agreed to elect Qin Chenfei as Chairman of the 11th Board of Directors and Chairman of the
June 12, 2026 at 9:55 AM
babygoldie.bsky.social
[1/4] 2026-06-12 17:45:21 - [Taiji Industrial: Elected Qin Chenfei as Chairman, who does not hold shares in the company] Taiji Industrial announced that the company held the second extraordinary shareholders meeting of 2026 on June 12, 2026, and reviewed and approved the by-election
June 12, 2026 at 9:55 AM
babygoldie.bsky.social
[4/4] of the company, and the company will handle industrial and commercial change registration. As of the announcement date, Qin Chenfei did not hold any shares in the company.
June 12, 2026 at 9:55 AM