Two-sided platforms are central to modern commerce and content sharing and often utilize A/B testing for developing new features. While user-side experiments are common, seller-side experiments become crucial for specific interventions and metrics. This paper investigates the effects of interference caused by feedback loops on seller-side experiments in two-sided platforms, with a particular focus on the counterfactual interleaving design, proposed in \citet{ha2020counterfactual,nandy2021b}. These feedback loops, often generated by pacing algorithms, cause outcomes from earlier sessions to influence subsequent ones. This paper contributes by creating a mathematical framework to analyze this interference, theoretically estimating its impact, and conducting empirical evaluations of the counterfactual interleaving design in real-world scenarios. Our research shows that feedback loops can result in misleading conclusions about the treatment effects.
The field of generative image inpainting and object insertion has made significant progress with the recent advent of latent diffusion models. Utilizing a precise object mask can greatly enhance these applications. However, due to the challenges users encounter in creating high-fidelity masks, there is a tendency for these methods to rely on more coarse masks (e.g., bounding box) for these applications. This results in limited control and compromised background content preservation. To overcome these limitations, we introduce SmartMask, which allows any novice user to create detailed masks for precise object insertion. Combined with a ControlNet-Inpaint model, our experiments demonstrate that SmartMask achieves superior object insertion quality, preserving the background content more effectively than previous methods. Notably, unlike prior works the proposed approach can also be used even without user-mask guidance, which allows it to perform mask-free object insertion at diverse positions and scales. Furthermore, we find that when used iteratively with a novel instruction-tuning based planning model, SmartMask can be used to design detailed layouts from scratch. As compared with user-scribble based layout design, we observe that SmartMask allows for better quality outputs with layout-to-image generation methods. Project page is available at https://smartmask-gen.github.io
Jointly considering multiple camera views (multi-view) is very effective for pedestrian detection under occlusion. For such multi-view systems, it is critical to have well-designed camera configurations, including camera locations, directions, and fields-of-view (FoVs). Usually, these configurations are crafted based on human experience or heuristics. In this work, we present a novel solution that features a transformer-based camera configuration generator. Using reinforcement learning, this generator autonomously explores vast combinations within the action space and searches for configurations that give the highest detection accuracy according to the training dataset. The generator learns advanced techniques like maximizing coverage, minimizing occlusion, and promoting collaboration. Across multiple simulation scenarios, the configurations generated by our transformer-based model consistently outperform random search, heuristic-based methods, and configurations designed by human experts, shedding light on future camera layout optimization.
In this paper, we present a comprehensive survey on online test-time adaptation (OTTA), a paradigm focused on adapting machine learning models to novel data distributions upon batch arrival. Despite the proliferation of OTTA methods recently, the field is mired in issues like ambiguous settings, antiquated backbones, and inconsistent hyperparameter tuning, obfuscating the real challenges and making reproducibility elusive. For clarity and a rigorous comparison, we classify OTTA techniques into three primary categories and subject them to benchmarks using the potent Vision Transformer (ViT) backbone to discover genuinely effective strategies. Our benchmarks span not only conventional corrupted datasets such as CIFAR-10/100-C and ImageNet-C but also real-world shifts embodied in CIFAR-10.1 and CIFAR-10-Warehouse, encapsulating variations across search engines and synthesized data by diffusion models. To gauge efficiency in online scenarios, we introduce novel evaluation metrics, inclusive of FLOPs, shedding light on the trade-offs between adaptation accuracy and computational overhead. Our findings diverge from existing literature, indicating: (1) transformers exhibit heightened resilience to diverse domain shifts, (2) the efficacy of many OTTA methods hinges on ample batch sizes, and (3) stability in optimization and resistance to perturbations are critical during adaptation, especially when the batch size is 1. Motivated by these insights, we pointed out promising directions for future research. The source code will be made available.
Molecular docking, a pivotal computational tool for drug discovery, predicts the binding interactions between small molecules (ligands) and target proteins (receptors). Conventional physics-based docking tools, though widely used, face limitations in precision due to restricted conformational sampling and imprecise scoring functions. Recent endeavors have employed deep learning techniques to enhance docking accuracy, but their generalization remains a concern due to limited training data. Leveraging the success of extensive and diverse data in other domains, we introduce HelixDock, a novel approach for site-specific molecular docking. Hundreds of millions of binding poses are generated by traditional docking tools, encompassing diverse protein targets and small molecules. Our deep learning-based docking model, a SE(3)-equivariant network, is pre-trained with this large-scale dataset and then fine-tuned with a small number of precise receptor-ligand complex structures. Comparative analyses against physics-based and deep learning-based baseline methods highlight HelixDock's superiority, especially on challenging test sets. Our study elucidates the scaling laws of the pre-trained molecular docking models, showcasing consistent improvements with increased model parameters and pre-train data quantities. Harnessing the power of extensive and diverse generated data holds promise for advancing AI-driven drug discovery.
In contrastive learning, two views of an original image generated by different augmentations are considered as a positive pair whose similarity is required to be high. Moreover, two views of two different images are considered as a negative pair, and their similarity is encouraged to be low. Normally, a single similarity measure given by a single projection head is used to evaluate positive and negative sample pairs, respectively. However, due to the various augmentation strategies and varying intra-sample similarity, augmented views from the same image are often not similar. Moreover, due to inter-sample similarity, augmented views of two different images may be more similar than augmented views from the same image. As such, enforcing a high similarity for positive pairs and a low similarity for negative pairs may not always be achievable, and in the case of some pairs, forcing so may be detrimental to the performance. To address this issue, we propose to use multiple projection heads, each producing a separate set of features. Our loss function for pre-training emerges from a solution to the maximum likelihood estimation over head-wise posterior distributions of positive samples given observations. The loss contains the similarity measure over positive and negative pairs, each re-weighted by an individual adaptive temperature that is regularized to prevent ill solutions. Our adaptive multi-head contrastive learning (AMCL) can be applied to and experimentally improves several popular contrastive learning methods such as SimCLR, MoCo and Barlow Twins. Such improvement is consistent under various backbones and linear probing epoches and is more significant when multiple augmentation methods are used.
Analyzing model performance in various unseen environments is a critical research problem in the machine learning community. To study this problem, it is important to construct a testbed with out-of-distribution test sets that have broad coverage of environmental discrepancies. However, existing testbeds typically either have a small number of domains or are synthesized by image corruptions, hindering algorithm design that demonstrates real-world effectiveness. In this paper, we introduce CIFAR-10-Warehouse, consisting of 180 datasets collected by prompting image search engines and diffusion models in various ways. Generally sized between 300 and 8,000 images, the datasets contain natural images, cartoons, certain colors, or objects that do not naturally appear. With CIFAR-10-W, we aim to enhance the evaluation and deepen the understanding of two generalization tasks: domain generalization and model accuracy prediction in various out-of-distribution environments. We conduct extensive benchmarking and comparison experiments and show that CIFAR-10-W offers new and interesting insights inherent to these tasks. We also discuss other fields that would benefit from CIFAR-10-W.
For object re-identification (re-ID), learning from synthetic data has become a promising strategy to cheaply acquire large-scale annotated datasets and effective models, with few privacy concerns. Many interesting research problems arise from this strategy, e.g., how to reduce the domain gap between synthetic source and real-world target. To facilitate developing more new approaches in learning from synthetic data, we introduce the Alice benchmarks, large-scale datasets providing benchmarks as well as evaluation protocols to the research community. Within the Alice benchmarks, two object re-ID tasks are offered: person and vehicle re-ID. We collected and annotated two challenging real-world target datasets: AlicePerson and AliceVehicle, captured under various illuminations, image resolutions, etc. As an important feature of our real target, the clusterability of its training set is not manually guaranteed to make it closer to a real domain adaptation test scenario. Correspondingly, we reuse existing PersonX and VehicleX as synthetic source domains. The primary goal is to train models from synthetic data that can work effectively in the real world. In this paper, we detail the settings of Alice benchmarks, provide an analysis of existing commonly-used domain adaptation methods, and discuss some interesting future directions. An online server will be set up for the community to evaluate methods conveniently and fairly.
Hand-crafted image quality metrics, such as PSNR and SSIM, are commonly used to evaluate model privacy risk under reconstruction attacks. Under these metrics, reconstructed images that are determined to resemble the original one generally indicate more privacy leakage. Images determined as overall dissimilar, on the other hand, indicate higher robustness against attack. However, there is no guarantee that these metrics well reflect human opinions, which, as a judgement for model privacy leakage, are more trustworthy. In this paper, we comprehensively study the faithfulness of these hand-crafted metrics to human perception of privacy information from the reconstructed images. On 5 datasets ranging from natural images, faces, to fine-grained classes, we use 4 existing attack methods to reconstruct images from many different classification models and, for each reconstructed image, we ask multiple human annotators to assess whether this image is recognizable. Our studies reveal that the hand-crafted metrics only have a weak correlation with the human evaluation of privacy leakage and that even these metrics themselves often contradict each other. These observations suggest risks of current metrics in the community. To address this potential risk, we propose a learning-based measure called SemSim to evaluate the Semantic Similarity between the original and reconstructed images. SemSim is trained with a standard triplet loss, using an original image as an anchor, one of its recognizable reconstructed images as a positive sample, and an unrecognizable one as a negative. By training on human annotations, SemSim exhibits a greater reflection of privacy leakage on the semantic level. We show that SemSim has a significantly higher correlation with human judgment compared with existing metrics. Moreover, this strong correlation generalizes to unseen datasets, models and attack methods.