How Well Do Feature-Additive Explainers Explain Feature-Additive Predictors?

Surging interest in deep learning from high-stakes domains has precipitated concern over the inscrutable nature of black box neural networks. Explainable AI (XAI) research has led to an abundance of explanation algorithms for these black boxes. Such post hoc explainers produce human-comprehensible explanations, however, their fidelity with respect to the model is not well understood - explanation evaluation remains one of the most challenging issues in XAI. In this paper, we ask a targeted but important question: can popular feature-additive explainers (e.g., LIME, SHAP, SHAPR, MAPLE, and PDP) explain feature-additive predictors? Herein, we evaluate such explainers on ground truth that is analytically derived from the additive structure of a model. We demonstrate the efficacy of our approach in understanding these explainers applied to symbolic expressions, neural networks, and generalized additive models on thousands of synthetic and several real-world tasks. Our results suggest that all explainers eventually fail to correctly attribute the importance of features, especially when a decision-making process involves feature interactions.

HomOpt: A Homotopy-Based Hyperparameter Optimization Method

Machine learning has achieved remarkable success over the past couple of decades, often attributed to a combination of algorithmic innovations and the availability of high-quality data available at scale. However, a third critical component is the fine-tuning of hyperparameters, which plays a pivotal role in achieving optimal model performance. Despite its significance, hyperparameter optimization (HPO) remains a challenging task for several reasons. Many HPO techniques rely on naive search methods or assume that the loss function is smooth and continuous, which may not always be the case. Traditional methods, like grid search and Bayesian optimization, often struggle to quickly adapt and efficiently search the loss landscape. Grid search is computationally expensive, while Bayesian optimization can be slow to prime. Since the search space for HPO is frequently high-dimensional and non-convex, it is often challenging to efficiently find a global minimum. Moreover, optimal hyperparameters can be sensitive to the specific dataset or task, further complicating the search process. To address these issues, we propose a new hyperparameter optimization method, HomOpt, using a data-driven approach based on a generalized additive model (GAM) surrogate combined with homotopy optimization. This strategy augments established optimization methodologies to boost the performance and effectiveness of any given method with faster convergence to the optimum on continuous, discrete, and categorical domain spaces. We compare the effectiveness of HomOpt applied to multiple optimization techniques (e.g., Random Search, TPE, Bayes, and SMAC) showing improved objective performance on many standardized machine learning benchmarks and challenging open-set recognition tasks.

On the Effectiveness of Image Manipulation Detection in the Age of Social Media

Image manipulation detection algorithms designed to identify local anomalies often rely on the manipulated regions being ``sufficiently'' different from the rest of the non-tampered regions in the image. However, such anomalies might not be easily identifiable in high-quality manipulations, and their use is often based on the assumption that certain image phenomena are associated with the use of specific editing tools. This makes the task of manipulation detection hard in and of itself, with state-of-the-art detectors only being able to detect a limited number of manipulation types. More importantly, in cases where the anomaly assumption does not hold, the detection of false positives in otherwise non-manipulated images becomes a serious problem. To understand the current state of manipulation detection, we present an in-depth analysis of deep learning-based and learning-free methods, assessing their performance on different benchmark datasets containing tampered and non-tampered samples. We provide a comprehensive study of their suitability for detecting different manipulations as well as their robustness when presented with non-tampered data. Furthermore, we propose a novel deep learning-based pre-processing technique that accentuates the anomalies present in manipulated regions to make them more identifiable by a variety of manipulation detection methods. To this end, we introduce an anomaly enhancement loss that, when used with a residual architecture, improves the performance of different detection algorithms with a minimal introduction of false positives on the non-manipulated data. Lastly, we introduce an open-source manipulation detection toolkit comprising a number of standard detection algorithms.

Human Activity Recognition in an Open World

Managing novelty in perception-based human activity recognition (HAR) is critical in realistic settings to improve task performance over time and ensure solution generalization outside of prior seen samples. Novelty manifests in HAR as unseen samples, activities, objects, environments, and sensor changes, among other ways. Novelty may be task-relevant, such as a new class or new features, or task-irrelevant resulting in nuisance novelty, such as never before seen noise, blur, or distorted video recordings. To perform HAR optimally, algorithmic solutions must be tolerant to nuisance novelty, and learn over time in the face of novelty. This paper 1) formalizes the definition of novelty in HAR building upon the prior definition of novelty in classification tasks, 2) proposes an incremental open world learning (OWL) protocol and applies it to the Kinetics datasets to generate a new benchmark KOWL-718, 3) analyzes the performance of current state-of-the-art HAR models when novelty is introduced over time, 4) provides a containerized and packaged pipeline for reproducing the OWL protocol and for modifying for any future updates to Kinetics. The experimental analysis includes an ablation study of how the different models perform under various conditions as annotated by Kinetics-AVA. The protocol as an algorithm for reproducing experiments using the KOWL-718 benchmark will be publicly released with code and containers at https://github.com/prijatelj/human-activity-recognition-in-an-open-world. The code may be used to analyze different annotations and subsets of the Kinetics datasets in an incremental open world fashion, as well as be extended as further updates to Kinetics are released.

Using Human Perception to Regularize Transfer Learning

Recent trends in the machine learning community show that models with fidelity toward human perceptual measurements perform strongly on vision tasks. Likewise, human behavioral measurements have been used to regularize model performance. But can we transfer latent knowledge gained from this across different learning objectives? In this work, we introduce PERCEP-TL (Perceptual Transfer Learning), a methodology for improving transfer learning with the regularization power of psychophysical labels in models. We demonstrate which models are affected the most by perceptual transfer learning and find that models with high behavioral fidelity -- including vision transformers -- improve the most from this regularization by as much as 1.9\% Top@1 accuracy points. These findings suggest that biologically inspired learning agents can benefit from human behavioral measurements as regularizers and psychophysical learned representations can be transferred to independent evaluation tasks.

Perceptual-Score: A Psychophysical Measure for Assessing the Biological Plausibility of Visual Recognition Models

For the last decade, convolutional neural networks (CNNs) have vastly superseded their predecessors in nearly all vision tasks in artificial intelligence, including object recognition. However, in spite of abundant advancements, they continue to pale in comparison to biological vision. This chasm has prompted the development of biologically-inspired models that have attempted to mimic the human visual system, primarily at a neural-level, which are evaluated using standard dataset benchmarks. However, more work is needed to understand how these models actually perceive the visual world. This article proposes a state-of-the-art procedure that generates a new metric, Perceptual-Score, which is grounded in visual psychophysics, and is capable of reliably estimating perceptual responses across numerous models -- representing a large range in complexity and biological inspiration. We perform the procedure on twelve models that vary in degree of biological inspiration and complexity, and compare the results against the aggregated results of 2,390 Amazon Mechanical Turk workers who together provided ~2.7 million perceptual responses. Each model's Perceptual-Score is compared against the state-of-the-art neural activity-based metric, Brain-Score. Our study indicates that models with high correlation to human perceptual behavior also have high correlation with the corresponding neural activity.

Guiding Machine Perception with Psychophysics

{G}{ustav} Fechner's 1860 delineation of psychophysics, the measurement of sensation in relation to its stimulus, is widely considered to be the advent of modern psychological science. In psychophysics, a researcher parametrically varies some aspects of a stimulus, and measures the resulting changes in a human subject's experience of that stimulus; doing so gives insight to the determining relationship between a sensation and the physical input that evoked it. This approach is used heavily in perceptual domains, including signal detection, threshold measurement, and ideal observer analysis. Scientific fields like vision science have always leaned heavily on the methods and procedures of psychophysics, but there is now growing appreciation of them by machine learning researchers, sparked by widening overlap between biological and artificial perception \cite{rojas2011automatic, scheirer2014perceptual,escalera2014chalearn,zhang2018agil, grieggs2021measuring}. Machine perception that is guided by behavioral measurements, as opposed to guidance restricted to arbitrarily assigned human labels, has significant potential to fuel further progress in artificial intelligence.

A Study of the Human Perception of Synthetic Faces

Advances in face synthesis have raised alarms about the deceptive use of synthetic faces. Can synthetic identities be effectively used to fool human observers? In this paper, we introduce a study of the human perception of synthetic faces generated using different strategies including a state-of-the-art deep learning-based GAN model. This is the first rigorous study of the effectiveness of synthetic face generation techniques grounded in experimental techniques from psychology. We answer important questions such as how often do GAN-based and more traditional image processing-based techniques confuse human observers, and are there subtle cues within a synthetic face image that cause humans to perceive it as a fake without having to search for obvious clues? To answer these questions, we conducted a series of large-scale crowdsourced behavioral experiments with different sources of face imagery. Results show that humans are unable to distinguish synthetic faces from real faces under several different circumstances. This finding has serious implications for many different applications where face images are presented to human users.

On the Objective Evaluation of Post Hoc Explainers

Many applications of data-driven models demand transparency of decisions, especially in health care, criminal justice, and other high-stakes environments. Modern trends in machine learning research have led to algorithms that are increasingly intricate to the degree that they are considered to be black boxes. In an effort to reduce the opacity of decisions, methods have been proposed to construe the inner workings of such models in a human-comprehensible manner. These post hoc techniques are described as being universal explainers - capable of faithfully augmenting decisions with algorithmic insight. Unfortunately, there is little agreement about what constitutes a "good" explanation. Moreover, current methods of explanation evaluation are derived from either subjective or proxy means. In this work, we propose a framework for the evaluation of post hoc explainers on ground truth that is directly derived from the additive structure of a model. We demonstrate the efficacy of the framework in understanding explainers by evaluating popular explainers on thousands of synthetic and several real-world tasks. The framework unveils that explanations may be accurate but misattribute the importance of individual features.

Handwriting Recognition with Novelty

This paper introduces an agent-centric approach to handle novelty in the visual recognition domain of handwriting recognition (HWR). An ideal transcription agent would rival or surpass human perception, being able to recognize known and new characters in an image, and detect any stylistic changes that may occur within or across documents. A key confound is the presence of novelty, which has continued to stymie even the best machine learning-based algorithms for these tasks. In handwritten documents, novelty can be a change in writer, character attributes, writing attributes, or overall document appearance, among other things. Instead of looking at each aspect independently, we suggest that an integrated agent that can process known characters and novelties simultaneously is a better strategy. This paper formalizes the domain of handwriting recognition with novelty, describes a baseline agent, introduces an evaluation protocol with benchmark data, and provides experimentation to set the state-of-the-art. Results show feasibility for the agent-centric approach, but more work is needed to approach human-levels of reading ability, giving the HWR community a formal basis to build upon as they solve this challenging problem.