Assessing Graphical Perception of Image Embedding Models using Channel Effectiveness

Recent advancements in vision models have greatly improved their ability to handle complex chart understanding tasks, like chart captioning and question answering. However, it remains challenging to assess how these models process charts. Existing benchmarks only roughly evaluate model performance without evaluating the underlying mechanisms, such as how models extract image embeddings. This limits our understanding of the model's ability to perceive fundamental graphical components. To address this, we introduce a novel evaluation framework to assess the graphical perception of image embedding models. For chart comprehension, we examine two main aspects of channel effectiveness: accuracy and discriminability of various visual channels. Channel accuracy is assessed through the linearity of embeddings, measuring how well the perceived magnitude aligns with the size of the stimulus. Discriminability is evaluated based on the distances between embeddings, indicating their distinctness. Our experiments with the CLIP model show that it perceives channel accuracy differently from humans and shows unique discriminability in channels like length, tilt, and curvature. We aim to develop this work into a broader benchmark for reliable visual encoders, enhancing models for precise chart comprehension and human-like perception in future applications.

Extracting Human Attention through Crowdsourced Patch Labeling

In image classification, a significant problem arises from bias in the datasets. When it contains only specific types of images, the classifier begins to rely on shortcuts - simplistic and erroneous rules for decision-making. This leads to high performance on the training dataset but inferior results on new, varied images, as the classifier's generalization capability is reduced. For example, if the images labeled as mustache consist solely of male figures, the model may inadvertently learn to classify images by gender rather than the presence of a mustache. One approach to mitigate such biases is to direct the model's attention toward the target object's location, usually marked using bounding boxes or polygons for annotation. However, collecting such annotations requires substantial time and human effort. Therefore, we propose a novel patch-labeling method that integrates AI assistance with crowdsourcing to capture human attention from images, which can be a viable solution for mitigating bias. Our method consists of two steps. First, we extract the approximate location of a target using a pre-trained saliency detection model supplemented by human verification for accuracy. Then, we determine the human-attentive area in the image by iteratively dividing the image into smaller patches and employing crowdsourcing to ascertain whether each patch can be classified as the target object. We demonstrated the effectiveness of our method in mitigating bias through improved classification accuracy and the refined focus of the model. Also, crowdsourced experiments validate that our method collects human annotation up to 3.4 times faster than annotating object locations with polygons, significantly reducing the need for human resources. We conclude the paper by discussing the advantages of our method in a crowdsourcing context, mainly focusing on aspects of human errors and accessibility.

Computational Approaches for App-to-App Retrieval and Design Consistency Check

Extracting semantic representations from mobile user interfaces (UI) and using the representations for designers' decision-making processes have shown the potential to be effective computational design support tools. Current approaches rely on machine learning models trained on small-sized mobile UI datasets to extract semantic vectors and use screenshot-to-screenshot comparison to retrieve similar-looking UIs given query screenshots. However, the usability of these methods is limited because they are often not open-sourced and have complex training pipelines for practitioners to follow, and are unable to perform screenshot set-to-set (i.e., app-to-app) retrieval. To this end, we (1) employ visual models trained with large web-scale images and test whether they could extract a UI representation in a zero-shot way and outperform existing specialized models, and (2) use mathematically founded methods to enable app-to-app retrieval and design consistency analysis. Our experiments show that our methods not only improve upon previous retrieval models but also enable multiple new applications.

Sanity Check for External Clustering Validation Benchmarks using Internal Validation Measures

We address the lack of reliability in benchmarking clustering techniques based on labeled datasets. A standard scheme in external clustering validation is to use class labels as ground truth clusters, based on the assumption that each class forms a single, clearly separated cluster. However, as such cluster-label matching (CLM) assumption often breaks, the lack of conducting a sanity check for the CLM of benchmark datasets casts doubt on the validity of external validations. Still, evaluating the degree of CLM is challenging. For example, internal clustering validation measures can be used to quantify CLM within the same dataset to evaluate its different clusterings but are not designed to compare clusterings of different datasets. In this work, we propose a principled way to generate between-dataset internal measures that enable the comparison of CLM across datasets. We first determine four axioms for between-dataset internal measures, complementing Ackerman and Ben-David's within-dataset axioms. We then propose processes to generalize internal measures to fulfill these new axioms, and use them to extend the widely used Calinski-Harabasz index for between-dataset CLM evaluation. Through quantitative experiments, we (1) verify the validity and necessity of the generalization processes and (2) show that the proposed between-dataset Calinski-Harabasz index accurately evaluates CLM across datasets. Finally, we demonstrate the importance of evaluating CLM of benchmark datasets before conducting external validation.