Robinson Piramuthu

Vishnu Sashank Dorbala, Gunnar Sigurdsson, Robinson Piramuthu et al. · uw

Household environments are visually diverse. Embodied agents performing Vision-and-Language Navigation (VLN) in the wild must be able to handle this diversity, while also following arbitrary language instructions. Recently, Vision-Language models like CLIP have shown great performance on the task of zero-shot object recognition. In this work, we ask if these models are also capable of zero-shot language grounding. In particular, we utilize CLIP to tackle the novel problem of zero-shot VLN using natural language referring expressions that describe target objects, in contrast to past work that used simple language templates describing object classes. We examine CLIP's capability in making sequential navigational decisions without any dataset-specific finetuning, and study how it influences the path that an agent takes. Our results on the coarse-grained instruction following task of REVERIE demonstrate the navigational capability of CLIP, surpassing the supervised baseline in terms of both success rate (SR) and success weighted by path length (SPL). More importantly, we quantitatively show that our CLIP-based zero-shot approach generalizes better to show consistent performance across environments when compared to SOTA, fully supervised learning approaches when evaluated via Relative Change in Success (RCS).

Amazon AGI, Aaron Langford, Aayush Shah et al. · amazon-science

We present Amazon Nova, a new generation of state-of-the-art foundation models that deliver frontier intelligence and industry-leading price performance. Amazon Nova Pro is a highly-capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks. Amazon Nova Lite is a low-cost multimodal model that is lightning fast for processing images, video, documents and text. Amazon Nova Micro is a text-only model that delivers our lowest-latency responses at very low cost. Amazon Nova Canvas is an image generation model that creates professional grade images with rich customization controls. Amazon Nova Reel is a video generation model offering high-quality outputs, customization, and motion control. Our models were built responsibly and with a commitment to customer trust, security, and reliability. We report benchmarking results for core capabilities, agentic performance, long context, functional adaptation, runtime performance, and human evaluation.

Gunnar A. Sigurdsson, Jesse Thomason, Gaurav S. Sukhatme et al. · uw

Household robots operate in the same space for years. Such robots incrementally build dynamic maps that can be used for tasks requiring remote object localization. However, benchmarks in robot learning often test generalization through inference on tasks in unobserved environments. In an observed environment, locating an object is reduced to choosing from among all object proposals in the environment, which may number in the 100,000s. Armed with this intuition, using only a generic vision-language scoring model with minor modifications for 3d encoding and operating in an embodied environment, we demonstrate an absolute performance gain of 9.84% on remote object grounding above state of the art models for REVERIE and of 5.04% on FAO. When allowed to pre-explore an environment, we also exceed the previous state of the art pre-exploration method on REVERIE. Additionally, we demonstrate our model on a real-world TurtleBot platform, highlighting the simplicity and usefulness of the approach. Our analysis outlines a "bag of tricks" essential for accomplishing this task, from utilizing 3d coordinates and context, to generalizing vision-language models to large 3d search spaces.

Brandon Trabucco, Gunnar Sigurdsson, Robinson Piramuthu et al.

Physically rearranging objects is an important capability for embodied agents. Visual room rearrangement evaluates an agent's ability to rearrange objects in a room to a desired goal based solely on visual input. We propose a simple yet effective method for this problem: (1) search for and map which objects need to be rearranged, and (2) rearrange each object until the task is complete. Our approach consists of an off-the-shelf semantic segmentation model, voxel-based semantic map, and semantic search policy to efficiently find objects that need to be rearranged. On the AI2-THOR Rearrangement Challenge, our method improves on current state-of-the-art end-to-end reinforcement learning-based methods that learn visual rearrangement policies from 0.53% correct rearrangement to 16.56%, using only 2.7% as many samples from the environment.

Siddharth Singi, Zhanpeng He, Alvin Pan et al.

In a Human-in-the-Loop paradigm, a robotic agent is able to act mostly autonomously in solving a task, but can request help from an external expert when needed. However, knowing when to request such assistance is critical: too few requests can lead to the robot making mistakes, but too many requests can overload the expert. In this paper, we present a Reinforcement Learning based approach to this problem, where a semi-autonomous agent asks for external assistance when it has low confidence in the eventual success of the task. The confidence level is computed by estimating the variance of the return from the current state. We show that this estimate can be iteratively improved during training using a Bellman-like recursion. On discrete navigation problems with both fully- and partially-observable state information, we show that our method makes effective use of a limited budget of expert calls at run-time, despite having no access to the expert at training time.

Shiyuan Huang, Robinson Piramuthu, Shih-Fu Chang et al.

In Video Question Answering (VideoQA), answering general questions about a video requires its visual information. Yet, video often contains redundant information irrelevant to the VideoQA task. For example, if the task is only to answer questions similar to "Is someone laughing in the video?", then all other information can be discarded. This paper investigates how many bits are really needed from the video in order to do VideoQA by introducing a novel Few-Bit VideoQA problem, where the goal is to accomplish VideoQA with few bits of video information (e.g., 10 bits). We propose a simple yet effective task-specific feature compression approach to solve this problem. Specifically, we insert a lightweight Feature Compression Module (FeatComp) into a VideoQA model which learns to extract task-specific tiny features as little as 10 bits, which are optimal for answering certain types of questions. We demonstrate more than 100,000-fold storage efficiency over MPEG4-encoded videos and 1,000-fold over regular floating point features, with just 2.0-6.6% absolute loss in accuracy, which is a surprising and novel finding. Finally, we analyze what the learned tiny features capture and demonstrate that they have eliminated most of the non-task-specific information, and introduce a Bit Activation Map to visualize what information is being stored. This decreases the privacy risk of data by providing k-anonymity and robustness to feature-inversion techniques, which can influence the machine learning community, allowing us to store data with privacy guarantees while still performing the task effectively.

Shiyuan Huang, Robinson Piramuthu, Vicente Ordonez et al.

In Video Question Answering, videos are often processed as a full-length sequence of frames to ensure minimal loss of information. Recent works have demonstrated evidence that sparse video inputs are sufficient to maintain high performance. However, they usually discuss the case of single frame selection. In our work, we extend the setting to multiple number of inputs and other modalities. We characterize the task with different input sparsity and provide a tool for doing that. Specifically, we use a Gumbel-based learnable selection module to adaptively select the best inputs for the final task. In this way, we experiment over public VideoQA benchmarks and provide analysis on how sparsified inputs affect the performance. From our experiments, we have observed only 5.2%-5.8% loss of performance with only 10% of video lengths, which corresponds to 2-4 frames selected from each video. Meanwhile, we also observed the complimentary behaviour between visual and textual inputs, even under highly sparsified settings, suggesting the potential of improving data efficiency for video-and-language tasks.

Jacob Zhiyuan Fang, Skyler Zheng, Vasu Sharma et al.

To build scalable models for challenging real-world tasks, it is important to learn from diverse, multi-modal data in various forms (e.g., videos, text, and images). Among the existing works, a plethora of them have focused on leveraging large but cumbersome cross-modal architectures. Regardless of their effectiveness, larger architectures unavoidably prevent the models from being extended to real-world applications, so building a lightweight VL architecture and an efficient learning schema is of great practical value. In this paper, we propose an Efficient Video-Language Model (dubbed as E-ViLM) and a masked video modeling (MVM) schema, assisted with a semantic vector-quantized tokenizer. In particular, our E-ViLM learns to reconstruct the semantic labels of masked video regions, produced by the pre-trained vector-quantized tokenizer, which discretizes the continuous visual signals into labels. We show that with our simple MVM task and regular VL pre-training modelings, our E-ViLM, despite its compactness, is able to learn expressive representations from Video-Language corpus and generalize well to extensive Video-Language tasks including video question answering, text-to-video retrieval, etc. In particular, our E-ViLM obtains obvious efficiency improvements by reaching competing performances with faster inference speed, i.e., our model reaches $39.3$% Top-$1$ accuracy on the MSRVTT benchmark, retaining $91.4$% of the accuracy of state-of-the-art larger VL architecture with only $15%$ parameters and $94.8%$ fewer GFLOPs. We also provide extensive ablative studies that validate the effectiveness of our proposed learning schema for E-ViLM.

Zhuoming Liu, Xiaofeng Gao, Feiyang Niu et al.

Chart understanding is crucial for deploying multimodal large language models (MLLMs) in real-world scenarios such as analyzing scientific papers and technical reports. Unlike natural images, charts pair a structured visual layout (spatial property) with an underlying data representation (textual property) -- grasping both is essential for precise, fine-grained chart reasoning. Motivated by this observation, we propose START, the Spatial and Textual learning for chART understanding. Specifically, we introduce (i) chart-element grounding and (ii) chart-to-code generation to strengthen an MLLM's understanding of both chart visual layout and data details. To facilitate spatial and textual learning, we propose the START-Dataset generated with a novel data-generation pipeline that first leverages an MLLM to translate real chart images into executable chart code, recovering the underlying data representation while preserving the visual distribution of real-world charts. We then evolve the code with a Large Language Model (LLM) to ascertain the positions of chart elements that capture the chart's visual structure, addressing challenges that existing methods cannot handle. To evaluate a model's ability to understand chart spatial structures, we propose the Chart Spatial understanding Benchmark (CS-Bench), filling a critical gap in comprehensive chart understanding evaluation. Leveraging spatial and textual learning, START delivers consistent gains across model sizes and benchmarks over the base models and surpasses prior state-of-the-art by a clear margin. Code, data and models will be publicly available.

Rohan Wadhawan, Fabrice Y Harel-Canada, Zi-Yi Dou et al.

Preference finetuning methods like Direct Preference Optimization (DPO) with AI-generated feedback have shown promise in aligning Large Vision-Language Models (LVLMs) with human preferences. However, existing techniques overlook the prevalence of noise in synthetic preference annotations in the form of stylistic and length biases. To this end, we introduce a hard-negative response generation framework based on LLM-guided response editing, that produces rejected responses with targeted errors, maintaining stylistic and length similarity to the accepted ones. Using this framework, we develop the VaPR dataset, comprising 30K high-quality samples, to finetune three LVLM families: LLaVA-V1.5, Qwen2VL & Qwen2.5VL (2B-13B sizes). Our VaPR models deliver significant performance improvements across ten benchmarks, achieving average gains of 6.5% (LLaVA), 4.0% (Qwen2VL), and 1.5% (Qwen2.5VL), with notable improvements on reasoning tasks. A scaling analysis shows that performance consistently improves with data size, with LLaVA models benefiting even at smaller scales. Moreover, VaPR reduces the tendency to answer "Yes" in binary questions - addressing a common failure mode in LVLMs like LLaVA. Lastly, we show that the framework generalizes to open-source LLMs as editors, with models trained on VaPR-OS achieving ~99% of the performance of models trained on \name, which is synthesized using GPT-4o. Our data, models, and code can be found on the project page https://vap-r.github.io

Junyi Zhang, Jia-Chen Gu, Wenbo Hu et al.

Existing medical reasoning benchmarks for vision-language models primarily focus on analyzing a patient's condition based on an image from a single visit. However, this setting deviates significantly from real-world clinical practice, where doctors typically refer to a patient's historical conditions to provide a comprehensive assessment by tracking their changes over time. In this paper, we introduce TemMed-Bench, the first benchmark designed for analyzing changes in patients' conditions between different clinical visits, which challenges large vision-language models (LVLMs) to reason over temporal medical images. TemMed-Bench consists of a test set comprising three tasks - visual question-answering (VQA), report generation, and image-pair selection - and a supplementary knowledge corpus of over 17,000 instances. With TemMed-Bench, we conduct an evaluation of six proprietary and six open-source LVLMs. Our results show that most LVLMs lack the ability to analyze patients' condition changes over temporal medical images, and a large proportion perform only at a random-guessing level in the closed-book setting. In contrast, GPT o3, o4-mini and Claude 3.5 Sonnet demonstrate comparatively decent performance, though they have yet to reach the desired level. Furthermore, we explore augmenting the input with both retrieved visual and textual modalities in the medical domain. We also show that multi-modal retrieval augmentation yields notably higher performance gains than no retrieval and textual retrieval alone across most models on our benchmark, with the VQA task showing an average improvement of 2.59%. Overall, we compose a benchmark grounded on real-world clinical practice, and it reveals LVLMs' limitations in temporal medical image reasoning, as well as highlighting the use of multi-modal retrieval augmentation as a potentially promising direction worth exploring to address this challenge.

Muratcan Cicek, Jinrong Xie, Qiaosong Wang et al.

Shopping is difficult for people with motor impairments. This includes online shopping. Proprietary software can emulate mouse and keyboard via head tracking. However, such a solution is not common for smartphones. Unlike desktop and laptop computers, they are also much easier to carry indoors and outdoors.To address this, we implement and open source button that is sensitive to head movements tracked from the front camera of iPhone X. This allows developers to integrate in eCommerce applications easily without requiring specialized knowledge. Other applications include gaming and use in hands-free situations such as during cooking, auto-repair. We built a sample online shopping application that allows users to easily browse between items from various categories and take relevant action just by head movements. We present results of user studies on this sample application and also include sensitivity studies based on two independent tests performed at 3 different distances to the screen.

Shuai Zheng, Fan Yang, M. Hadi Kiapour et al.

Understanding clothes from a single image has strong commercial and cultural impacts on modern societies. However, this task remains a challenging computer vision problem due to wide variations in the appearance, style, brand and layering of clothing items. We present a new database called ModaNet, a large-scale collection of images based on Paperdoll dataset. Our dataset provides 55,176 street images, fully annotated with polygons on top of the 1 million weakly annotated street images in Paperdoll. ModaNet aims to provide a technical benchmark to fairly evaluate the progress of applying the latest computer vision techniques that rely on large data for fashion understanding. The rich annotation of the dataset allows to measure the performance of state-of-the-art algorithms for object detection, semantic segmentation and polygon prediction on street fashion images in detail. The polygon-based annotation dataset has been released https://github.com/eBay/modanet, we also host the leaderboard at EvalAI: https://evalai.cloudcv.org/featured-challenges/136/overview.

James F. Mullen, Prasoon Goyal, Robinson Piramuthu et al.

Home robots intend to make their users lives easier. Our work assists in this goal by enabling robots to inform their users of dangerous or unsanitary anomalies in their home. Some examples of these anomalies include the user leaving their milk out, forgetting to turn off the stove, or leaving poison accessible to children. To move towards enabling home robots with these abilities, we have created a new dataset, which we call SafetyDetect. The SafetyDetect dataset consists of 1000 anomalous home scenes, each of which contains unsafe or unsanitary situations for an agent to detect. Our approach utilizes large language models (LLMs) alongside both a graph representation of the scene and the relationships between the objects in the scene. Our key insight is that this connected scene graph and the object relationships it encodes enables the LLM to better reason about the scene -- especially as it relates to detecting dangerous or unsanitary situations. Our most promising approach utilizes GPT-4 and pursues a categorization technique where object relations from the scene graph are classified as normal, dangerous, unsanitary, or dangerous for children. This method is able to correctly identify over 90% of anomalous scenarios in the SafetyDetect Dataset. Additionally, we conduct real world experiments on a ClearPath TurtleBot where we generate a scene graph from visuals of the real world scene, and run our approach with no modification. This setup resulted in little performance loss. The SafetyDetect Dataset and code will be released to the public upon this papers publication.

Brandon Trabucco, Gunnar Sigurdsson, Robinson Piramuthu et al.

The predominant approach for training web navigation agents is to gather human demonstrations for a set of popular websites and hand-written tasks, but it is becoming clear that human data is an inefficient resource. We develop a pipeline to facilitate internet-scale training for agents without laborious human annotations. In the first stage, an LLM annotates 150k sites with agentic tasks. In the next stage, LLM agents complete tasks and produce trajectories. In the final stage, an LLM filters trajectories by judging their success. Language models are powerful data curation tools, identifying harmful content with an accuracy of 97%, judging successful trajectories with an accuracy of 82.6%, and producing effective data. We train agents based on Qwen 3 1.7B that are competitive with frontier LLMs as web agents, while being smaller and faster. Our top agent reaches a success rate of 56.9%, outperforming the data collection policy Qwen 3 235B, a 235 times larger Llama 4 Maverick, and reaching 94.7% of the performance of Gemini 2.5 Flash. We are releasing code, models and data at: https://data-for-agents.github.io.

Xuehai He, Jian Zheng, Jacob Zhiyuan Fang et al.

Controllable text-to-image (T2I) diffusion models generate images conditioned on both text prompts and semantic inputs of other modalities like edge maps. Nevertheless, current controllable T2I methods commonly face challenges related to efficiency and faithfulness, especially when conditioning on multiple inputs from either the same or diverse modalities. In this paper, we propose a novel Flexible and Efficient method, FlexEControl, for controllable T2I generation. At the core of FlexEControl is a unique weight decomposition strategy, which allows for streamlined integration of various input types. This approach not only enhances the faithfulness of the generated image to the control, but also significantly reduces the computational overhead typically associated with multimodal conditioning. Our approach achieves a reduction of 41% in trainable parameters and 30% in memory usage compared with Uni-ControlNet. Moreover, it doubles data efficiency and can flexibly generate images under the guidance of multiple input conditions of various modalities.

Vishnu Sashank Dorbala, Prasoon Goyal, Robinson Piramuthu et al.

We present and tackle the problem of Embodied Question Answering (EQA) with Situational Queries (S-EQA) in a household environment. Unlike prior EQA work tackling simple queries that directly reference target objects and properties ("What is the color of the car?"), situational queries (such as "Is the house ready for sleeptime?") are challenging as they require the agent to correctly identify multiple object-states (Doors: Closed, Lights: Off, etc.) and reach a consensus on their states for an answer. Towards this objective, we first introduce a novel Prompt-Generate-Evaluate (PGE) scheme that wraps around an LLM's output to generate unique situational queries and corresponding consensus object information. PGE is used to generate 2K datapoints in the VirtualHome simulator, which is then annotated for ground truth answers via a large scale user-study conducted on M-Turk. With a high rate of answerability (97.26%) on this study, we establish that LLMs are good at generating situational data. However, in evaluating the data using an LLM, we observe a low correlation of 46.2% with the ground truth human annotations; indicating that while LLMs are good at generating situational data, they struggle to answer them according to consensus. When asked for reasoning, we observe the LLM often goes against commonsense in justifying its answer. Finally, we utilize PGE to generate situational data in a real-world environment, exposing LLM hallucination in generating reliable object-states when a structured scene graph is unavailable. To the best of our knowledge, this is the first work to introduce EQA in the context of situational queries and also the first to present a generative approach for query creation. We aim to foster research on improving the real-world usability of embodied agents through this work.

Yinhong Liu, Jianfeng He, Hang Su et al.

Multimodal Dialogue Summarization (MDS) is a critical task with wide-ranging applications. To support the development of effective MDS models, robust automatic evaluation methods are essential for reducing both cost and human effort. However, such methods require a strong meta-evaluation benchmark grounded in human annotations. In this work, we introduce MDSEval, the first meta-evaluation benchmark for MDS, consisting image-sharing dialogues, corresponding summaries, and human judgments across eight well-defined quality aspects. To ensure data quality and richfulness, we propose a novel filtering framework leveraging Mutually Exclusive Key Information (MEKI) across modalities. Our work is the first to identify and formalize key evaluation dimensions specific to MDS. We benchmark state-of-the-art modal evaluation methods, revealing their limitations in distinguishing summaries from advanced MLLMs and their susceptibility to various bias.

Aishwarya Padmakumar, Jesse Thomason, Ayush Shrivastava et al.

Robots operating in human spaces must be able to engage in natural language interaction with people, both understanding and executing instructions, and using conversation to resolve ambiguity and recover from mistakes. To study this, we introduce TEACh, a dataset of over 3,000 human--human, interactive dialogues to complete household tasks in simulation. A Commander with access to oracle information about a task communicates in natural language with a Follower. The Follower navigates through and interacts with the environment to complete tasks varying in complexity from "Make Coffee" to "Prepare Breakfast", asking questions and getting additional information from the Commander. We propose three benchmarks using TEACh to study embodied intelligence challenges, and we evaluate initial models' abilities in dialogue understanding, language grounding, and task execution.

Ayush Shrivastava, Karthik Gopalakrishnan, Yang Liu et al.

Interactive robots navigating photo-realistic environments need to be trained to effectively leverage and handle the dynamic nature of dialogue in addition to the challenges underlying vision-and-language navigation (VLN). In this paper, we present VISITRON, a multi-modal Transformer-based navigator better suited to the interactive regime inherent to Cooperative Vision-and-Dialog Navigation (CVDN). VISITRON is trained to: i) identify and associate object-level concepts and semantics between the environment and dialogue history, ii) identify when to interact vs. navigate via imitation learning of a binary classification head. We perform extensive pre-training and fine-tuning ablations with VISITRON to gain empirical insights and improve performance on CVDN. VISITRON's ability to identify when to interact leads to a natural generalization of the game-play mode introduced by Roman et al. (arXiv:2005.00728) for enabling the use of such models in different environments. VISITRON is competitive with models on the static CVDN leaderboard and attains state-of-the-art performance on the Success weighted by Path Length (SPL) metric.

Yun-Chun Chen, Marco Piccirilli, Robinson Piramuthu et al.

We consider the task of estimating 3D human pose and shape from videos. While existing frame-based approaches have made significant progress, these methods are independently applied to each image, thereby often leading to inconsistent predictions. In this work, we present a video-based learning algorithm for 3D human pose and shape estimation. The key insights of our method are two-fold. First, to address the inconsistent temporal prediction issue, we exploit temporal information in videos and propose a self-attention module that jointly considers short-range and long-range dependencies across frames, resulting in temporally coherent estimations. Second, we model human motion with a forecasting module that allows the transition between adjacent frames to be smooth. We evaluate our method on the 3DPW, MPI-INF-3DHP, and Human3.6M datasets. Extensive experimental results show that our algorithm performs favorably against the state-of-the-art methods.

Yu-Ting Chang, Qiaosong Wang, Wei-Chih Hung et al.

Obtaining object response maps is one important step to achieve weakly-supervised semantic segmentation using image-level labels. However, existing methods rely on the classification task, which could result in a response map only attending on discriminative object regions as the network does not need to see the entire object for optimizing the classification loss. To tackle this issue, we propose a principled and end-to-end train-able framework to allow the network to pay attention to other parts of the object, while producing a more complete and uniform response map. Specifically, we introduce the mixup data augmentation scheme into the classification network and design two uncertainty regularization terms to better interact with the mixup strategy. In experiments, we conduct extensive analysis to demonstrate the proposed method and show favorable performance against state-of-the-art approaches.

Yu-Ting Chang, Qiaosong Wang, Wei-Chih Hung et al.

Existing weakly-supervised semantic segmentation methods using image-level annotations typically rely on initial responses to locate object regions. However, such response maps generated by the classification network usually focus on discriminative object parts, due to the fact that the network does not need the entire object for optimizing the objective function. To enforce the network to pay attention to other parts of an object, we propose a simple yet effective approach that introduces a self-supervised task by exploiting the sub-category information. Specifically, we perform clustering on image features to generate pseudo sub-categories labels within each annotated parent class, and construct a sub-category objective to assign the network to a more challenging task. By iteratively clustering image features, the training process does not limit itself to the most discriminative object parts, hence improving the quality of the response maps. We conduct extensive analysis to validate the proposed method and show that our approach performs favorably against the state-of-the-art approaches.

Xiao Ma, Lina Mezghani, Kimberly Wilber et al.

As any savvy online shopper knows, second-hand peer-to-peer marketplaces are filled with images of mixed quality. How does image quality impact marketplace outcomes, and can quality be automatically predicted? In this work, we conducted a large-scale study on the quality of user-generated images in peer-to-peer marketplaces. By gathering a dataset of common second-hand products (~75,000 images) and annotating a subset with human-labeled quality judgments, we were able to model and predict image quality with decent accuracy (~87%). We then conducted two studies focused on understanding the relationship between these image quality scores and two marketplace outcomes: sales and perceived trustworthiness. We show that image quality is associated with higher likelihood that an item will be sold, though other factors such as view count were better predictors of sales. Nonetheless, we show that high quality user-generated images selected by our models outperform stock imagery in eliciting perceptions of trust from users. Our findings can inform the design of future marketplaces and guide potential sellers to take better product images.

M. Hadi Kiapour, Robinson Piramuthu

While lots of people may think branding begins and ends with a logo, fashion brands communicate their uniqueness through a wide range of visual cues such as color, patterns and shapes. In this work, we analyze learned visual representations by deep networks that are trained to recognize fashion brands. In particular, the activation strength and extent of neurons are studied to provide interesting insights about visual brand expressions. The proposed method identifies where a brand stands in the spectrum of branding strategy, i.e., from trademark-emblazoned goods with bold logos to implicit no logo marketing. By quantifying attention maps, we are able to interpret the visual characteristics of a brand present in a single image and model the general design direction of a brand as a whole. We further investigate versatility of neurons and discover "specialists" that are highly brand-specific and "generalists" that detect diverse visual features. A human experiment based on three main visual scenarios of fashion brands is conducted to verify the alignment of our quantitative measures with the human perception of brands. This paper demonstrate how deep networks go beyond logos in order to recognize clothing brands in an image.

Bryan A. Plummer, M. Hadi Kiapour, Shuai Zheng et al.

In this paper, we introduce an attribute-based interactive image search which can leverage human-in-the-loop feedback to iteratively refine image search results. We study active image search where human feedback is solicited exclusively in visual form, without using relative attribute annotations used by prior work which are not typically found in many datasets. In order to optimize the image selection strategy, a deep reinforcement model is trained to learn what images are informative rather than rely on hand-crafted measures typically leveraged in prior work. Additionally, we extend the recently introduced Conditional Similarity Network to incorporate global similarity in training visual embeddings, which results in more natural transitions as the user explores the learned similarity embeddings. Our experiments demonstrate the effectiveness of our approach, producing compelling results on both active image search and image attribute representation tasks.

Kevin Lin, Fan Yang, Qiaosong Wang et al.

Fine-grained image search is still a challenging problem due to the difficulty in capturing subtle differences regardless of pose variations of objects from fine-grained categories. In practice, a dynamic inventory with new fine-grained categories adds another dimension to this challenge. In this work, we propose an end-to-end network, called FGGAN, that learns discriminative representations by implicitly learning a geometric transformation from multi-view images for fine-grained image search. We integrate a generative adversarial network (GAN) that can automatically handle complex view and pose variations by converting them to a canonical view without any predefined transformations. Moreover, in an open-set scenario, our network is able to better match images from unseen and unknown fine-grained categories. Extensive experiments on two public datasets and a newly collected dataset have demonstrated the outstanding robust performance of the proposed FGGAN in both closed-set and open-set scenarios, providing as much as 10% relative improvement compared to baselines.

Bryan A. Plummer, Paige Kordas, M. Hadi Kiapour et al.

This paper presents an approach for grounding phrases in images which jointly learns multiple text-conditioned embeddings in a single end-to-end model. In order to differentiate text phrases into semantically distinct subspaces, we propose a concept weight branch that automatically assigns phrases to embeddings, whereas prior works predefine such assignments. Our proposed solution simplifies the representation requirements for individual embeddings and allows the underrepresented concepts to take advantage of the shared representations before feeding them into concept-specific layers. Comprehensive experiments verify the effectiveness of our approach across three phrase grounding datasets, Flickr30K Entities, ReferIt Game, and Visual Genome, where we obtain a (resp.) 4%, 3%, and 4% improvement in grounding performance over a strong region-phrase embedding baseline.

Mahyar Najibi, Fan Yang, Qiaosong Wang et al.

In this work, we propose an efficient and effective approach for unconstrained salient object detection in images using deep convolutional neural networks. Instead of generating thousands of candidate bounding boxes and refining them, our network directly learns to generate the saliency map containing the exact number of salient objects. During training, we convert the ground-truth rectangular boxes to Gaussian distributions that better capture the ROI regarding individual salient objects. During inference, the network predicts Gaussian distributions centered at salient objects with an appropriate covariance, from which bounding boxes are easily inferred. Notably, our network performs saliency map prediction without pixel-level annotations, salient object detection without object proposals, and salient object subitizing simultaneously, all in a single pass within a unified framework. Extensive experiments show that our approach outperforms existing methods on various datasets by a large margin, and achieves more than 100 fps with VGG16 network on a single GPU during inference.

Fan Yang, Ajinkya Kale, Yury Bubnov et al.

In this paper, we propose a novel end-to-end approach for scalable visual search infrastructure. We discuss the challenges we faced for a massive volatile inventory like at eBay and present our solution to overcome those. We harness the availability of large image collection of eBay listings and state-of-the-art deep learning techniques to perform visual search at scale. Supervised approach for optimized search limited to top predicted categories and also for compact binary signature are key to scale up without compromising accuracy and precision. Both use a common deep neural network requiring only a single forward inference. The system architecture is presented with in-depth discussions of its basic components and optimizations for a trade-off between search relevance and latency. This solution is currently deployed in a distributed cloud infrastructure and fuels visual search in eBay ShopBot and Close5. We show benchmark on ImageNet dataset on which our approach is faster and more accurate than several unsupervised baselines. We share our learnings with the hope that visual search becomes a first class citizen for all large scale search engines rather than an afterthought.

Bolei Zhou, Vignesh Jagadeesh, Robinson Piramuthu

Discovering visual knowledge from weakly labeled data is crucial to scale up computer vision recognition system, since it is expensive to obtain fully labeled data for a large number of concept categories. In this paper, we propose ConceptLearner, which is a scalable approach to discover visual concepts from weakly labeled image collections. Thousands of visual concept detectors are learned automatically, without human in the loop for additional annotation. We show that these learned detectors could be applied to recognize concepts at image-level and to detect concepts at image region-level accurately. Under domain-specific supervision, we further evaluate the learned concepts for scene recognition on SUN database and for object detection on Pascal VOC 2007. ConceptLearner shows promising performance compared to fully supervised and weakly supervised methods.

Kota Hara, Vignesh Jagadeesh, Robinson Piramuthu

In this work, we propose and address a new computer vision task, which we call fashion item detection, where the aim is to detect various fashion items a person in the image is wearing or carrying. The types of fashion items we consider in this work include hat, glasses, bag, pants, shoes and so on. The detection of fashion items can be an important first step of various e-commerce applications for fashion industry. Our method is based on state-of-the-art object detection method pipeline which combines object proposal methods with a Deep Convolutional Neural Network. Since the locations of fashion items are in strong correlation with the locations of body joints positions, we incorporate contextual information from body poses in order to improve the detection performance. Through the experiments, we demonstrate the effectiveness of the proposed method.

Kevin Shih, Wei Di, Vignesh Jagadeesh et al.

Text is ubiquitous in the artificial world and easily attainable when it comes to book title and author names. Using the images from the book cover set from the Stanford Mobile Visual Search dataset and additional book covers and metadata from openlibrary.org, we construct a large scale book cover retrieval dataset, complete with 100K distractor covers and title and author strings for each. Because our query images are poorly conditioned for clean text extraction, we propose a method for extracting a matching noisy and erroneous OCR readings and matching it against clean author and book title strings in a standard document look-up problem setup. Finally, we demonstrate how to use this text-matching as a feature in conjunction with popular retrieval features such as VLAD using a simple learning setup to achieve significant improvements in retrieval accuracy over that of either VLAD or the text alone.

Qiaosong Wang, Vignesh Jagadeesh, Bryan Ressler et al.

Recent advances in consumer depth sensors have created many opportunities for human body measurement and modeling. Estimation of 3D body shape is particularly useful for fashion e-commerce applications such as virtual try-on or fit personalization. In this paper, we propose a method for capturing accurate human body shape and anthropometrics from a single consumer grade depth sensor. We first generate a large dataset of synthetic 3D human body models using real-world body size distributions. Next, we estimate key body measurements from a single monocular depth image. We combine body measurement estimates with local geometry features around key joint positions to form a robust multi-dimensional feature vector. This allows us to conduct a fast nearest-neighbor search to every sample in the dataset and return the closest one. Compared to existing methods, our approach is able to predict accurate full body parameters from a partial view using measurement parameters learned from the synthetic dataset. Furthermore, our system is capable of generating 3D human mesh models in real-time, which is significantly faster than methods which attempt to model shape and pose deformations. To validate the efficiency and applicability of our system, we collected a dataset that contains frontal and back scans of 83 clothed people with ground truth height and weight. Experiments on real-world dataset show that the proposed method can achieve real-time performance with competing results achieving an average error of 1.9 cm in estimated measurements.

Zhicheng Yan, Hao Zhang, Robinson Piramuthu et al.

In image classification, visual separability between different object categories is highly uneven, and some categories are more difficult to distinguish than others. Such difficult categories demand more dedicated classifiers. However, existing deep convolutional neural networks (CNN) are trained as flat N-way classifiers, and few efforts have been made to leverage the hierarchical structure of categories. In this paper, we introduce hierarchical deep CNNs (HD-CNNs) by embedding deep CNNs into a category hierarchy. An HD-CNN separates easy classes using a coarse category classifier while distinguishing difficult classes using fine category classifiers. During HD-CNN training, component-wise pretraining is followed by global finetuning with a multinomial logistic loss regularized by a coarse category consistency term. In addition, conditional executions of fine category classifiers and layer parameter compression make HD-CNNs scalable for large-scale visual recognition. We achieve state-of-the-art results on both CIFAR100 and large-scale ImageNet 1000-class benchmark datasets. In our experiments, we build up three different HD-CNNs and they lower the top-1 error of the standard CNNs by 2.65%, 3.1% and 1.1%, respectively.

Wei Di, Anurag Bhardwaj, Vignesh Jagadeesh et al.

Fashion, and especially apparel, is the fastest-growing category in online shopping. As consumers requires sensory experience especially for apparel goods for which their appearance matters most, images play a key role not only in conveying crucial information that is hard to express in text, but also in affecting consumer's attitude and emotion towards the product. However, research related to e-commerce product image has mostly focused on quality at perceptual level, but not the quality of content, and the way of presenting. This study aims to address the effectiveness of types of image in showcasing fashion apparel in terms of its attractiveness, i.e. the ability to draw consumer's attention, interest, and in return their engagement. We apply advanced vision technique to quantize attractiveness using three common display types in fashion filed, i.e. human model, mannequin, and flat. We perform two-stage study by starting with large scale behavior data from real online market, then moving to well designed user experiment to further deepen our understandings on consumer's reasoning logic behind the action. We propose a Fisher noncentral hypergeometric distribution based user choice model to quantitatively evaluate user's preference. Further, we investigate the potentials to leverage visual impact for a better search that caters to user's preference. A visual attractiveness based re-ranking model that incorporates both presentation efficacy and user preference is proposed. We show quantitative improvement by promoting visual attractiveness into search on top of relevance.

Anurag Bhardwaj, Vignesh Jagadeesh, Wei Di et al.

We describe a completely automated large scale visual recommendation system for fashion. Existing approaches have primarily relied on purely computational models to solving this problem that ignore the role of users in the system. In this paper, we propose to overcome this limitation by incorporating a user-centric design of visual fashion recommendations. Specifically, we propose a technique that augments 'user preferences' in models by exploiting elasticity in fashion choices. We further design a user study on these choices and gather results from the 'wisdom of crowd' for deeper analysis. Our key insights learnt through these results suggest that fashion preferences when constrained to a particular class, contain important behavioral signals that are often ignored in recommendation design. Further, presence of such classes also reflect strong correlations to visual perception which can be utilized to provide aesthetically pleasing user experiences. Finally, we illustrate that user approval of visual fashion recommendations can be substantially improved by carefully incorporating these user-centric feedback into the system framework.

Zixuan Wang, Wei Di, Anurag Bhardwaj et al.

We present a novel compact image descriptor for large scale image search. Our proposed descriptor - Geometric VLAD (gVLAD) is an extension of VLAD (Vector of Locally Aggregated Descriptors) that incorporates weak geometry information into the VLAD framework. The proposed geometry cues are derived as a membership function over keypoint angles which contain evident and informative information but yet often discarded. A principled technique for learning the membership function by clustering angles is also presented. Further, to address the overhead of iterative codebook training over real-time datasets, a novel codebook adaptation strategy is outlined. Finally, we demonstrate the efficacy of proposed gVLAD based retrieval framework where we achieve more than 15% improvement in mAP over existing benchmarks.

Vignesh Jagadeesh, Robinson Piramuthu, Anurag Bhardwaj et al.

We describe a completely automated large scale visual recommendation system for fashion. Our focus is to efficiently harness the availability of large quantities of online fashion images and their rich meta-data. Specifically, we propose four data driven models in the form of Complementary Nearest Neighbor Consensus, Gaussian Mixture Models, Texture Agnostic Retrieval and Markov Chain LDA for solving this problem. We analyze relative merits and pitfalls of these algorithms through extensive experimentation on a large-scale data set and baseline them against existing ideas from color science. We also illustrate key fashion insights learned through these experiments and show how they can be employed to design better recommendation systems. Finally, we also outline a large-scale annotated data set of fashion images (Fashion-136K) that can be exploited for future vision research.