Solving complex visual tasks such as “Who invented the musical instrument on the right?” involves a composition of skills: understanding space, recognizing instruments, and also retrieving prior knowledge. Recent work shows promise by decomposing such tasks using a large language model (LLM) into an executable program that invokes specialized vision models. However, generated programs are error-prone: they omit necessary steps, include spurious ones, and are unable to recover when the specialized models give incor-rect outputs. Moreover, they require loading multiple models, incurring high latency and computation costs. We propose Visual Program Distillation (VPD), an instruction tuning framework that produces a vision-language model (VLM) ca-pable of solving complex visual tasks with a single forward pass. VPD distills the reasoning ability of LLMs by using them to sample multiple candidate programs, which are then executed and verified to identify a correct one. It translates each correct program into a language description of the reasoning steps, which are then distilled into a VLM. Exten-sive experiments show that VPD improves the VLM's ability to count, understand spatial relations, and reason compositionally. Our VPD-trained PaLI-X outperforms all prior VLMs, achieving state-of-the-art performance across complex vision tasks, including MMBench, OK-VQA, A-OKVQA, TallyQA, POPE, and Hateful Memes. An evaluation with human annotators also confirms that VPD improves model response factuality and consistency. Finally, experiments on content moderation demonstrate that VPD is also helpful for adaptation to real-world applications with limited data.

The application of computer vision methods to nuanced, subjective concepts is growing. While crowdsourcing has served the vision community well for most objective tasks (such as labeling a "zebra"), it now falters on tasks where there is substantial subjectivity in the concept (such as identifying "gourmet tuna"). However, empowering any user to develop a classifier for their concept is technically difficult: users are neither machine learning experts nor have the patience to label thousands of examples. In reaction, we introduce the problem of Agile Modeling: the process of turning any subjective visual concept into a computer vision model through real-time user-in-the-loop interactions. We instantiate an Agile Modeling prototype for image classification and show through a user study (N=14) that users can create classifiers with minimal effort in under 30 minutes. We compare this user driven process with the traditional crowdsourcing paradigm and find that the crowd’s notion often differs from that of the user’s, especially as the concepts become more subjective. Finally, we scale our experiments with simulations of users training classifiers for ImageNet21k categories to further demonstrate the efficacy of the approach.

Computer vision models learn to perform a task by capturing relevant statistics from training data. It has been shown that models learn spurious age, gender, and race correlations when trained for seemingly unrelated tasks like activity recognition or image captioning. Various mitigation techniques have been presented to prevent models from utilizing or learning such biases. However, there has been little systematic comparison between these techniques. We design a simple but surprisingly effective visual recognition benchmark for studying bias mitigation. Using this benchmark, we provide a thorough analysis of a wide range of techniques. We highlight the shortcomings of popular adversarial training approaches for bias mitigation, propose a simple but similarly effective alternative to the inference-time Reducing Bias Amplification method of Zhao et al., and design a domain-independent training technique that outperforms all other methods. Finally, we validate our findings on the attribute classification task in the CelebA dataset, where attribute presence is known to be correlated with the gender of people in the image, and demonstrate that the proposed technique is effective at mitigating real-world gender bias.

Microtask crowdsourcing is increasingly critical to the creation of extremely large datasets. As a result, crowd workers spend weeks or months repeating the exact same tasks, making it necessary to understand their behavior over these long periods of time. We utilize three large, longitudinal datasets of nine million annotations collected from Amazon Mechanical Turk to examine claims that workers fatigue or satisfice over these long periods, producing lower quality work. We find that, contrary to these claims, workers are extremely stable in their quality over the entire period. To understand whether workers set their quality based on the task's requirements for acceptance, we then perform an experiment where we vary the required quality for a large crowdsourcing task. Workers did not adjust their quality based on the acceptance threshold: workers who were above the threshold continued working at their usual quality level, and workers below the threshold self-selected themselves out of the task. Capitalizing on this consistency, we demonstrate that it is possible to predict workers' long-term quality using just a glimpse of their quality on the first five tasks.

Most natural videos contain numerous events. For example, in a video of a “man playing a piano”, the video might also contain “another man dancing” or “a crowd clapping”. We introduce the task of dense-captioning events, which involves both detecting and describing events in a video. We propose a new model that is able to identify all events in a single pass of the video while simultaneously describing the detected events with natural language. Our model introduces a variant of an existing proposal module that is designed to capture both short as well as long events that span minutes. To capture the dependencies between the events in a video, our model introduces a new captioning module that uses contextual information from past and future events to jointly describe all events. We also introduce ActivityNet Captions, a large-scale benchmark for dense-captioning events. ActivityNet Captions contains 20k videos amounting to 849 video hours with 100k total descriptions, each with its unique start and end time. Finally, we report performances of our model for dense-captioning events, video retrieval and localization.

Despite progress in perceptual tasks such as image classification, computers still perform poorly on cognitive tasks such as image description and question answering. Cognition is core to tasks that involve not just recognizing, but reasoning about our visual world. However, models used to tackle the rich content in images for cognitive tasks are still being trained using the same datasets designed for perceptual tasks. To achieve success at cognitive tasks, models need to understand the interactions and relationships between objects in an image. When asked “What vehicle is the person riding?”, computers will need to identify the objects in an image as well as the relationships riding(man, carriage) and pulling(horse, carriage) to answer correctly that “the person is riding a horse-drawn carriage.” In this paper, we present the Visual Genome dataset to enable the modeling of such relationships. We collect dense annotations of objects, attributes, and relationships within each image to learn these models. Specifically, our dataset contains over 108K images where each image has an average of $$35$$ objects, $$26$$ attributes, and $$21$$ pairwise relationships between objects. We canonicalize the objects, attributes, relationships, and noun phrases in region descriptions and questions answer pairs to WordNet synsets. Together, these annotations represent the densest and largest dataset of image descriptions, objects, attributes, relationships, and question answer pairs.

Microtask crowdsourcing has enabled dataset advances in social science and machine learning, but existing crowdsourcing schemes are too expensive to scale up with the expanding volume of data. To scale and widen the applicability of crowdsourcing, we present a technique that produces extremely rapid judgments for binary and categorical labels. Rather than punishing all errors, which causes workers to proceed slowly and deliberately, our technique speeds up workers' judgments to the point where errors are acceptable and even expected. We demonstrate that it is possible to rectify these errors by randomizing task order and modeling response latency. We evaluate our technique on a breadth of common labeling tasks such as image verification, word similarity, sentiment analysis and topic classification. Where prior work typically achieves a 0.25x to 1x speedup over fixed majority vote, our approach often achieves an order of magnitude (10x) speedup.

A Haptic Jamming tactile display, consisting of an array of thin particle jamming cells, can change its shape and mechanical properties simultaneously through a combination of vacuuming individual cells, pressurizing the air chamber beneath the surface, and pinning the nodes between the cells at various heights. In previous particle jamming devices for haptics and soft robotics, shape has typically been commanded open-loop or manipulated directly by a human user. A new algorithm was designed for the three types of actuation inputs for a Haptic Jamming surface, using the depth map provided by an RGB-D sensor as shape feedback for closed-loop control to match a desired surface input. To test the closed-loop control accuracy of the system, a mass-spring model of the Haptic Jamming device generated three unique surface shapes as desired inputs into the controller with a mean height range of 25.1 mm. The average correlation coefficient between the desired input shape and the experimental output generated by the controller on the actual device across four trials for each shape was 0.88, with an average height error of 2.7 mm. When the desired input surface is generated from a 3D model of an object that a Haptic Jamming surface cannot necessarily re-create, the difference between the input and the output increases substantially. However, simulation of a larger array suggests that a Haptic Jamming surface can provide a compelling match for these more complicated shapes.