skip to main content


The NSF Public Access Repository (NSF-PAR) system and access will be unavailable from 11:00 PM ET on Friday, April 12 until 2:00 AM ET on Saturday, April 13 due to maintenance. We apologize for the inconvenience.

Title: From heatmaps to structured explanations of image classifiers

This paper summarizes our endeavors in the past few years in terms of explaining image classifiers, with the aim of including negative results and insights we have gained. The paper starts with describing the explainable neural network (XNN), which attempts to extract and visualize several high‐level concepts purely from the deep network, without relying on human linguistic concepts. This helps users understand network classifications that are less intuitive and substantially improves user performance on a difficult fine‐grained classification task of discriminating among different species of seagulls. Realizing that an important missing piece is a reliable heatmap visualization tool, we have developed integrated‐gradient optimized saliency (I‐GOS) and iGOS++ utilizing integrated gradients to avoid local optima in heatmap generation, which improved the performance across all resolutions. During the development of those visualizations, we realized that for a significant number of images, the classifier has multiple different paths to reach a confident prediction. This has led to our recent development of structured attention graphs, an approach that utilizes beam search to locate multiple coarse heatmaps for a single image, and compactly visualizes a set of heatmaps by capturing how different combinations of image regions impact the confidence of a classifier. Through the research process, we have learned much about insights in building deep network explanations, the existence and frequency of multiple explanations, and various tricks of the trade that make explanations work. In this paper, we attempt to share those insights and opinions with the readers with the hope that some of them will be informative for future researchers on explainable deep learning.

more » « less
Author(s) / Creator(s):
 ;  ;  ;  ;  ;  ;  
Publisher / Repository:
Wiley Blackwell (John Wiley & Sons)
Date Published:
Journal Name:
Applied AI Letters
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. End-to-end deep learning models are increasingly applied to safety-critical human activity recognition (HAR) applications, e.g., healthcare monitoring and smart home control, to reduce developer burden and increase the performance and robustness of prediction models. However, integrating HAR models in safety-critical applications requires trust, and recent approaches have aimed to balance the performance of deep learning models with explainable decision-making for complex activity recognition. Prior works have exploited the compositionality of complex HAR (i.e., higher-level activities composed of lower-level activities) to form models with symbolic interfaces, such as concept-bottleneck architectures, that facilitate inherently interpretable models. However, feature engineering for symbolic concepts-as well as the relationship between the concepts-requires precise annotation of lower-level activities by domain experts, usually with fixed time windows, all of which induce a heavy and error-prone workload on the domain expert. In this paper, we introduce X-CHAR, an eXplainable Complex Human Activity Recognition model that doesn't require precise annotation of low-level activities, offers explanations in the form of human-understandable, high-level concepts, while maintaining the robust performance of end-to-end deep learning models for time series data. X-CHAR learns to model complex activity recognition in the form of a sequence of concepts. For each classification, X-CHAR outputs a sequence of concepts and a counterfactual example as the explanation. We show that the sequence information of the concepts can be modeled using Connectionist Temporal Classification (CTC) loss without having accurate start and end times of low-level annotations in the training dataset-significantly reducing developer burden. We evaluate our model on several complex activity datasets and demonstrate that our model offers explanations without compromising the prediction accuracy in comparison to baseline models. Finally, we conducted a mechanical Turk study to show that the explanations provided by our model are more understandable than the explanations from existing methods for complex activity recognition. 
    more » « less
  2. The local explanation provides heatmaps on images to explain how Convolutional Neural Networks (CNNs) derive their output. Due to its visual straightforwardness, the method has been one of the most popular explainable AI (XAI) methods for diagnosing CNNs. Through our formative study (S1), however, we captured ML engineers' ambivalent perspective about the local explanation as a valuable and indispensable envision in building CNNs versus the process that exhausts them due to the heuristic nature of detecting vulnerability. Moreover, steering the CNNs based on the vulnerability learned from the diagnosis seemed highly challenging. To mitigate the gap, we designed DeepFuse, the first interactive design that realizes the direct feedback loop between a user and CNNs in diagnosing and revising CNN's vulnerability using local explanations. DeepFuse helps CNN engineers to systemically search unreasonable local explanations and annotate the new boundaries for those identified as unreasonable in a labor-efficient manner. Next, it steers the model based on the given annotation such that the model doesn't introduce similar mistakes. We conducted a two-day study (S2) with 12 experienced CNN engineers. Using DeepFuse, participants made a more accurate and reasonable model than the current state-of-the-art. Also, participants found the way DeepFuse guides case-based reasoning can practically improve their current practice. We provide implications for design that explain how future HCI-driven design can move our practice forward to make XAI-driven insights more actionable.

    more » « less
  3. Heatmap regression-based models have significantly advanced the progress of facial landmark detection. However, the lack of structural constraints always generates inaccurate heatmaps resulting in poor landmark detection performance. While hierarchical structure modeling methods have been proposed to tackle this issue, they all heavily rely on manually designed tree structures. The designed hierarchical structure is likely to be completely corrupted due to the missing or inaccurate prediction of landmarks. To the best of our knowledge, in the context of deep learning, no work before has investigated how to automatically model proper structures for facial landmarks, by discovering their inherent relations. In this paper, we propose a novel Hierarchical Structured Landmark Ensemble (HSLE) model for learning robust facial landmark detection, by using it as the structural constraints. Different from existing approaches of manually designing structures, our proposed HSLE model is constructed automatically via discovering the most robust patterns so HSLE has the ability to robustly depict both local and holistic landmark structures simultaneously. Our proposed HSLE can be readily plugged into any existing facial landmark detection baselines for further performance improvement. Extensive experimental results demonstrate our approach significantly outperforms the baseline by a large margin to achieve a state-of-the-art performance. 
    more » « less
  4. Abstract Despite the increasingly successful application of neural networks to many problems in the geosciences, their complex and nonlinear structure makes the interpretation of their predictions difficult, which limits model trust and does not allow scientists to gain physical insights about the problem at hand. Many different methods have been introduced in the emerging field of eXplainable Artificial Intelligence (XAI), which aims at attributing the network’s prediction to specific features in the input domain. XAI methods are usually assessed by using benchmark datasets (such as MNIST or ImageNet for image classification). However, an objective, theoretically derived ground truth for the attribution is lacking for most of these datasets, making the assessment of XAI in many cases subjective. Also, benchmark datasets specifically designed for problems in geosciences are rare. Here, we provide a framework, based on the use of additively separable functions, to generate attribution benchmark datasets for regression problems for which the ground truth of the attribution is known a priori. We generate a large benchmark dataset and train a fully connected network to learn the underlying function that was used for simulation. We then compare estimated heatmaps from different XAI methods to the ground truth in order to identify examples where specific XAI methods perform well or poorly. We believe that attribution benchmarks as the ones introduced herein are of great importance for further application of neural networks in the geosciences, and for more objective assessment and accurate implementation of XAI methods, which will increase model trust and assist in discovering new science. 
    more » « less
  5. Medical image segmentation is one of the most challenging tasks in medical image analysis and widely developed for many clinical applications. While deep learning-based approaches have achieved impressive performance in semantic segmentation, they are limited to pixel-wise settings with imbalanced-class data problems and weak boundary object segmentation in medical images. In this paper, we tackle those limitations by developing a new two-branch deep network architecture which takes both higher level features and lower level features into account. The first branch extracts higher level feature as region information by a common encoder-decoder network structure such as Unet and FCN, whereas the second branch focuses on lower level features as support information around the boundary and processes in parallel to the first branch. Our key contribution is the second branch named Narrow Band Active Contour (NB-AC) attention model which treats the object contour as a hyperplane and all data inside a narrow band as support information that influences the position and orientation of the hyperplane. Our proposed NB-AC attention model incorporates the contour length with the region energy involving a fixed-width band around the curve or surface. The proposed network loss contains two fitting terms: (i) a high level feature (i.e., region) fitting term from the first branch; (ii) a lower level feature (i.e., contour) fitting term from the second branch including the (ii1) length of the object contour and (ii2) regional energy functional formed by the homogeneity criterion of both the inner band and outer band neighboring the evolving curve or surface. The proposed NB-AC loss can be incorporated into both 2D and 3D deep network architectures. The proposed network has been evaluated on different challenging medical image datasets, including DRIVE, iSeg17, MRBrainS18 and Brats18. The experimental results have shown that the proposed NB-AC loss outperforms other mainstream loss functions: Cross Entropy, Dice, Focal on two common segmentation frameworks Unet and FCN. Our 3D network which is built upon the proposed NB-AC loss and 3DUnet framework achieved state-of-the-art results on multiple volumetric datasets. 
    more » « less