High-quality, pixel-level annotated datasets are crucial for training deep learning models, while their creation is often labor-intensive, time-consuming, and costly. Generative diffusion models have then gained prominence for producing synthetic datasets, yet existing text-to-data methods struggle with generating complex scenes involving multiple objects and intricate spatial arrangements. To address these limitations, we introduce FlexDataset, a framework that pioneers the composition-to-data (C2D) paradigm. FlexDataset generates high-fidelity synthetic datasets with versatile annotations, tailored for tasks like salient object detection, depth estimation, and segmentation. Leveraging a meticulously designed composition-to-image (C2I) framework, it offers precise positional and categorical control. Our Versatile Annotation Generation (VAG) Plan A further enhances efficiency by exploiting rich latent representations through tuned perception decoders, reducing annotation time by nearly fivefold. FlexDataset allows unlimited generation of customized, multi-instance and multi-category (MIMC) annotated data. Extensive experiments show that FlexDataset sets a new standard in synthetic dataset generation across multiple datasets and tasks, including zero-shot and long-tail scenarios.

The widespread adoption of Batch Normalization (BN) in contemporary deep neural architectures has demonstrated significant efficacy, particularly in the domain of Unsupervised Domain Adaptation (UDA) for cross-domain applications. Notwithstanding its success, extant BN variants often conflate source and target domain information within identical channels, potentially compromising transferability due to inter-domain feature misalignment. To address this limitation, we introduce Refined Batch Normalization (RBN), a novel normalization paradigm that leverages estimated shift to quantify discrepancies between estimated population statistics and their expected values. Our pivotal observation reveals that estimated shift can accumulate through BN stacking within the network, potentially degrading target domain performance. We elucidate how RBN mitigates this accumulation, thereby enhancing overall system efficacy. The practical implementation of this technique is realized through the RBNBlock, which supplants conventional BN with RBN in the bottleneck architecture of residual networks. Extensive empirical evaluation across diverse cross-domain benchmarks corroborates the superiority of RBN in augmenting inter-domain transferability. This perspective transcends immediate performance metrics, offering a foundational lens through which subsequent research can more deeply understand and refine the interplay between normalization strategies and domain adaptation.