Gang Ren

Xingxuan Zhang, Gang Ren, Han Yu et al.

We argue that progress toward general intelligence requires complementary foundation models grounded in language, the physical world, and structured data. This report presents LimiX-16M and LimiX-2M, two instantiations of our large structured-data models (LDMs). Both models treat structured data as a joint distribution over variables and missingness, thus capable of addressing a wide range of tabular tasks through query-based conditional prediction via a single model. They are pretrained using masked joint-distribution modeling with an episodic, context-conditional objective, supporting rapid, training-free adaptation at inference. We evaluate LimiX models across 11 large structured-data benchmarks with broad regimes of sample size, feature dimensionality, class number, categorical-to-numerical feature ratio, missingness, and sample-to-feature ratios. LimiX-16M consistently surpasses strong baselines, as shown in Figure 1 and Figure 2. The superiority holds across a wide range of tasks, such as classification, regression, missing value imputation, and data generation, often by substantial margins, while avoiding task-specific architectures or bespoke training per task. Notably, LimiX-2M delivers strong results under tight compute and memory budgets. We also present the first scaling law study for LDMs, revealing how data and model scaling jointly influence downstream performance and offering quantitative guidance for tabular foundation modeling. All LimiX models are publicly accessible under Apache 2.0.

Gang Ren, Wenbin Li, Eamonn O'Neill

As interactive devices become pervasive, people are beginning to looking for more advanced interaction with televisions in the living room. Interactive television has the potential to offer a very engaging experience. But most common user tasks are still challenging with such systems, such as menu selection or text input. And little work has been done on understanding and sup-porting the effective design of freehand interaction with an TV in the living room. In this paper, we perform two studies investi-gating freehand gestural interaction with a consumer level sensor, which is suitable for TV scenarios. In the first study, we inves-tigate a range of design factors for tiled layout menu selection, including wearable feedback, push gesture depth, target size and position in motor space. The results show that tactile and audio feedback have no significant effect on performance and prefer-ence, and these results inform potential designs for high selection performance. In the second study, we investigate a common TV user task of text input using freehand gesture. We design and evaluate two virtual keyboard layouts and three freehand selec-tion methods. Results show that ease of use and error tolerance can be both achieved using a text entry method utilizing a dual circle layout and an expanding target selection technique. Finally, we propose design guidelines for effective, usable and com-fortable freehand gestural interaction for interactive TV based on the findings.

Wenbin Li, Yang Chen, JeeHang Lee et al.

It is hard to estimate optical flow given a realworld video sequence with camera shake and other motion blur. In this paper, we first investigate the blur parameterization for video footage using near linear motion elements. we then combine a commercial 3D pose sensor with an RGB camera, in order to film video footage of interest together with the camera motion. We illustrates that this additional camera motion/trajectory channel can be embedded into a hybrid framework by interleaving an iterative blind deconvolution and warping based optical flow scheme. Our method yields improved accuracy within three other state-of-the-art baselines given our proposed ground truth blurry sequences; and several other realworld sequences filmed by our imaging system.