arXiv subjects

Gordon Chen

Publications and source records attributed to Gordon Chen.

4 recordsLinked to original sources

arXiv2026

PSEC6: an 8-Channel 40 GSa/s Waveform Sampling ASIC in TSMC 65nm with 10.24 GHz PLL

Picosecond level timing resolution is a prerequisite capability for improved coincidence matching, time-of-flight measurements, and secondary vertex reconstruction. Here, we present the specification, design, and simulation results for a new Application Specific Integrated Circuit (ASIC), called PSEC6, in the TSMC 65nm process. It features 8 channels, a maximum sampling rate of 40 GSa/s, a buffer length of 204.8 nanoseconds, and a 10.24 GHz Phase Locked Loop (PLL), which is the first of its kind in the 65nm CMOS process. The event readout rate is 32 kHz, with the digitization done by an off-chip Analog-to-Digital Converter (ADC). Simulations predict a 4.0 GHz analog input bandwidth and 20 mW per channel during sampling; the 10.24 GHz PLL has a predicted jitter of 550 fs RMS at 15.7 mW. The paper describes the sampling architecture, chip signal paths, PLL design, and presents simulation results.

physics.ins-det↗

arXiv2026

Prompt Relay: Inference-Time Temporal Control for Multi-Event Video Generation

Video diffusion models have achieved remarkable progress in generating high-quality videos. However, these models struggle to represent the temporal succession of multiple events in real-world videos and lack explicit mechanisms to control when semantic concepts appear, how long they persist, and the order in which multiple events occur. Such control is especially important for movie-grade video synthesis, where coherent storytelling depends on precise timing, duration, and transitions between events. When using a single paragraph-style prompt to describe a sequence of complex events, models often exhibit semantic entanglement, where concepts intended for different moments in the video bleed into one another, resulting in poor text-video alignment. To address these limitations, we propose Prompt Relay, an inference-time, plug-and-play method to enable fine-grained temporal control in multi-event video generation, requiring no architectural modifications and no additional computational overhead. Prompt Relay introduces a penalty into the cross-attention mechanism, so that each temporal segment attends only to its assigned prompt, allowing the model to represent one semantic concept at a time and thereby improving temporal prompt alignment, reducing semantic interference, and enhancing visual quality.

cs.CV↗

arXiv2025

VChain: Chain-of-Visual-Thought for Reasoning in Video Generation

Recent video generation models can produce smooth and visually appealing clips, but they often struggle to synthesize complex dynamics with a coherent chain of consequences. Accurately modeling visual outcomes and state transitions over time remains a core challenge. In contrast, large language and multimodal models (e.g., GPT-4o) exhibit strong visual state reasoning and future prediction capabilities. To bridge these strengths, we introduce VChain, a novel inference-time chain-of-visual-thought framework that injects visual reasoning signals from multimodal models into video generation. Specifically, VChain contains a dedicated pipeline that leverages large multimodal models to generate a sparse set of critical keyframes as snapshots, which are then used to guide the sparse inference-time visual-state adaptation of a pre-trained video generator only at these key moments. Our approach is tuning-efficient, introduces minimal overhead and avoids dense supervision. Extensive experiments on complex, multi-step scenarios show that VChain significantly enhances the quality of generated videos.

cs.CV↗

arXiv2025

Stencil: Subject-Driven Generation with Context Guidance

Recent text-to-image diffusion models can generate striking visuals from text prompts, but they often fail to maintain subject consistency across generations and contexts. One major limitation of current fine-tuning approaches is the inherent trade-off between quality and efficiency. Fine-tuning large models improves fidelity but is computationally expensive, while fine-tuning lightweight models improves efficiency but compromises image fidelity. Moreover, fine-tuning pre-trained models on a small set of images of the subject can damage the existing priors, resulting in suboptimal results. To this end, we present Stencil, a novel framework that jointly employs two diffusion models during inference. Stencil efficiently fine-tunes a lightweight model on images of the subject, while a large frozen pre-trained model provides contextual guidance during inference, injecting rich priors to enhance generation with minimal overhead. Stencil excels at generating high-fidelity, novel renditions of the subject in less than a minute, delivering state-of-the-art performance and setting a new benchmark in subject-driven generation.

cs.CV↗