An official website of the United States government
Every artist has a creative process that draws inspiration from previous artists and their works. Today, “inspiration” has been automated by generative music models. The black box nature of these models obscures the identity of the works that influence their creative output. As a result, users may inadvertently appropriate, misuse, or copy existing artists’ works. We establish a replicable methodology to systematically identify similar pieces of music audio in a manner that is useful for understanding training data attribution. A key aspect of our approach is to harness an effective music audio similarity measure. We compare the effect of applying CLMR [50] and CLAP [55] embeddings to similarity measurement in a set of 5 million audio clips used to train VampNet [24], a recent open source generative music model. We validate this approach with a human listening study. We also explore the effect that modifications of an audio example (e.g., pitch shifting, time stretching, background noise) have on similarity measurements. This work is foundational to incorporating automated influence attribution into generative modeling, which promises to let model creators and users move from ignorant appropriation to informed creation. Audio samples that accompany this paper are available at https://tinyurl.com/exploring-musical-roots.
more »
« less
IdeateRelate: An Examples Gallery That Helps Creators Explore Ideas in Relation to Their Own
Creating truly original ideas requires extensive knowledge of existing ideas. Navigating prior examples can help people to understand what has already been done and to assess the quality of their own ideas through comparison. The creativity literature has suggested that the conceptual distance between a proposed solution and a potential inspiration can influence one's thinking. However, less is known about how creators might use data about conceptual distance when exploring a large repository of ideas. To investigate this, we created a novel tool for exploring examples called IdeateRelate that visualizes 600+ COVID-related ideas, organized by their similarity to a new idea. In an experiment that compared the IdeateRelate visualization to a simple list of examples, we found that users in the Viz condition leveraged both semantic and categorical similarity, curated a more similar set of examples, and adopted more language from examples into their iterated ideas (without negatively affecting the overall novelty). We discuss implications for creating adaptive interfaces that provide creative inspiration in response to designers' ideas throughout an iterative design process.
more »
« less
- Award ID(s):
- 1821618
- PAR ID:
- 10602792
- Publisher / Repository:
- Association for Computing Machinery (ACM)
- Date Published:
- Journal Name:
- Proceedings of the ACM on Human-Computer Interaction
- Volume:
- 5
- Issue:
- CSCW2
- ISSN:
- 2573-0142
- Format(s):
- Medium: X Size: p. 1-18
- Size(s):
- p. 1-18
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Computational modeling tools present unique opportunities and challenges for student learning. Each tool has a representational system that impacts the kinds of explorations students engage in. Inquiry aligned with a tool’s representational system can support more productive engagement toward target learning goals. However, little research has examined how teachers can make visible the ways students’ ideas about a phenomenon can be expressed and explored within a tool’s representational system. In this paper, we elaborate on the construct of ontological alignment—that is, identifying and leveraging points of resonance between students’ existing ideas and the representational system of a tool. Using interaction analysis, we identify alignment practices adopted by a science teacher and her students in a computational agent-based modeling unit. Specifically, we describe three practices: (1) Elevating student ideas relevant to the tool’s representational system; (2) Exploring and testing links between students’ conceptual and computational models; and (3) Drawing on evidence resonant with the tool’s representational system to differentiate between theories. Finally, we discuss the pedagogical value of ontological alignment as a way to leverage students’ ideas in alignment with a tool’s representational system and suggest the presented practices as exemplary ways to support students’ computational modeling for science learning.more » « less
-
Exposure to ideas in domains outside a scientist's own may benefit her in reformulating existing research problems in novel ways and discovering new application domains for existing solution ideas. While improved performance in scholarly search engines can help scientists efficiently identify relevant advances in domains they may already be familiar with, it may fall short of helping them explore diverse ideas \textit{outside} such domains. In this paper we explore the design of systems aimed at augmenting the end-user ability in cross-domain exploration with flexible query specification. To this end, we develop an exploratory search system in which end-users can select a portion of text core to their interest from a paper abstract and retrieve papers that have a high similarity to the user-selected core aspect but differ in terms of domains. Furthermore, end-users can `zoom in' to specific domain clusters to retrieve more papers from them and understand nuanced differences within the clusters. Our case studies with scientists uncover opportunities and design implications for systems aimed at facilitating cross-domain exploration and inspiration.more » « less
-
Assessing similarity between design ideas is an inherent part of many design evaluations to measure novelty. In such evaluation tasks, humans excel at making mental connections among diverse knowledge sets to score ideas on their uniqueness. However, their decisions about novelty are often subjective and difficult to explain. In this paper, we demonstrate a way to uncover human judgment of design idea similarity using two-dimensional (2D) idea maps. We derive these maps by asking participants for simple similarity comparisons of the form “Is idea A more similar to idea B or to idea C?” We show that these maps give insight into the relationships between ideas and help understand the design domain. We also propose that novel ideas can be identified by finding outliers on these idea maps. To demonstrate our method, we conduct experimental evaluations on two datasets—colored polygons (known answer) and milk frother sketches (unknown answer). We show that idea maps shed light on factors considered by participants in judging idea similarity and the maps are robust to noisy ratings. We also compare physical maps made by participants on a white-board to their computationally generated idea maps to compare how people think about spatial arrangement of design items. This method provides a new direction of research into deriving ground truth novelty metrics by combining human judgments and computational methods.more » « less
-
Deep learning models are vulnerable to adversarial examples. Most of current adversarial attacks add pixel-wise perturbations restricted to some L^p-norm, and defense models are evaluated also on adversarial examples restricted inside L^p-norm balls. However, we wish to explore adversarial examples exist beyond L^p-norm balls and their implications for attacks and defenses. In this paper, we focus on adversarial images generated by transformations. We start with color transformation and propose two gradient-based attacks. Since L^p-norm is inappropriate for measuring image quality in the transformation space, we use the similarity between transformations and the Structural Similarity Index. Next, we explore a larger transformation space consisting of combinations of color and affine transformations. We evaluate our transformation attacks on three data sets --- CIFAR10, SVHN, and ImageNet --- and their corresponding models. Finally, we perform retraining defenses to evaluate the strength of our attacks. The results show that transformation attacks are powerful. They find high-quality adversarial images that have higher transferability and misclassification rates than C&W's L^p attacks, especially at high confidence levels. They are also significantly harder to defend against by retraining than C&W's L^p attacks. More importantly, exploring different attack spaces makes it more challenging to train a universally robust model.more » « less
