LaSagnA: Language-based Segmentation Assistant for Complex Queries

Recent advancements have empowered Large Language Models for Vision (vLLMs) to generate detailed perceptual outcomes, including bounding boxes and masks. Nonetheless, there are two constraints that restrict the further application of these vLLMs: the incapability of handling multiple targets per query and the failure to identify the absence of query objects in the image. In this study, we acknowledge that the main cause of these problems is the insufficient complexity of training queries. Consequently, we define the general sequence format for complex queries. Then we incorporate a semantic segmentation task in the current pipeline to fulfill the requirements of training data. Furthermore, we present three novel strategies to effectively handle the challenges arising from the direct integration of the proposed format. The effectiveness of our model in processing complex queries is validated by the comparable results with conventional methods on both close-set and open-set semantic segmentation datasets. Additionally, we outperform a series of vLLMs in reasoning and referring segmentation, showcasing our model's remarkable capabilities. We release the code at https://github.com/congvvc/LaSagnA.

Mutually-aware Sub-Graphs Differentiable Architecture Search

Differentiable architecture search is prevalent in the field of NAS because of its simplicity and efficiency, where two paradigms, multi-path algorithms and single-path methods, are dominated. Multi-path framework (e.g. DARTS) is intuitive but suffers from memory usage and training collapse. Single-path methods (e.g.GDAS and ProxylessNAS) mitigate the memory issue and shrink the gap between searching and evaluation but sacrifice the performance. In this paper, we propose a conceptually simple yet efficient method to bridge these two paradigms, referred as Mutually-aware Sub-Graphs Differentiable Architecture Search (MSG-DAS). The core of our framework is a differentiable Gumbel-TopK sampler that produces multiple mutually exclusive single-path sub-graphs. To alleviate the severer skip-connect issue brought by multiple sub-graphs setting, we propose a Dropblock-Identity module to stabilize the optimization. To make best use of the available models (super-net and sub-graphs), we introduce a memory-efficient super-net guidance distillation to improve training. The proposed framework strikes a balance between flexible memory usage and searching quality. We demonstrate the effectiveness of our methods on ImageNet and CIFAR10, where the searched models show a comparable performance as the most recent approaches.

A Coordinate-wise Optimization Algorithm for Sparse Inverse Covariance Selection

Sparse inverse covariance selection is a fundamental problem for analyzing dependencies in high dimensional data. However, such a problem is difficult to solve since it is NP-hard. Existing solutions are primarily based on convex approximation and iterative hard thresholding, which only lead to sub-optimal solutions. In this work, we propose a coordinate-wise optimization algorithm to solve this problem which is guaranteed to converge to a coordinate-wise minimum point. The algorithm iteratively and greedily selects one variable or swaps two variables to identify the support set, and then solves a reduced convex optimization problem over the support set to achieve the greatest descent. As a side contribution of this paper, we propose a Newton-like algorithm to solve the reduced convex sub-problem, which is proven to always converge to the optimal solution with global linear convergence rate and local quadratic convergence rate. Finally, we demonstrate the efficacy of our method on synthetic data and real-world data sets. As a result, the proposed method consistently outperforms existing solutions in terms of accuracy.