Generative AI is reshaping fashion by enabling virtual looks and avatars making it essential to find real products that best match AI-generated styles. We propose an end-to-end product search system that has been deployed in a real-world, internet scale which ensures that AI-generated looks presented to users are matched with the most visually and semantically similar products from the indexed vector space. The search pipeline is composed of four key components: query generation, vectorization, candidate retrieval, and reranking based on AI-generated looks. Recommendation quality is evaluated using human-judged accuracy scores. The system currently serves more than 350,000 AI Looks in production per day, covering diverse product categories across global markets of over 12 million products. In our experiments, we observed that across multiple annotators and categories, CLIP outperformed alternative models by a small relative margin of 3--7\% in mean opinion scores. These improvements, though modest in absolute numbers, resulted in noticeably better user perception matches, establishing CLIP as the most reliable backbone for production deployment.
Adaptive patching is a recent and compelling proposal for time-series Transformers: allocate finer patches where the sequence looks locally informative. This paper asks under what conditions a content-adaptive patching operator should outperform a tuned uniform one. Local heterogeneity alone is not enough: under pointwise forecasting losses, a complex-looking region is not automatically one where finer patching reduces the loss. We model patching as a budgeted bitrate allocation and derive an explicit threshold that a dynamic patching rule must satisfy to beat a well-tuned uniform baseline, then bound the achievable improvement both locally (a quadratic surrogate) and globally (a strong-convexity bound under the model's assumptions). Two structural results follow: without a coupling constraint, scalar local complexity cannot produce a non-uniform optimum under a common loss landscape; and once the backbone is trained to its representation-aware optimum, the alignment gain collapses around a well-tuned uniform patch size. To test these predictions, we run a controlled isolation study on three representative architectures, replacing each adaptive mechanism with a uniform patch-size s
The growing interest in eXplainable Artificial Intelligence (XAI) has stimulated research on models with built-in interpretability, among which part-prototype models are particularly prominent. Part-Prototype Models (PPMs) classify inputs by comparing them to learned prototypes and provide human-understandable explanations of the form "this looks like that". Despite this intrinsic interpretability, PPMs have not yet emerged as a competitive alternative to post-hoc explanation methods. This survey reviews work published between 2019 and 2025 and derives a taxonomy of the challenges faced by current PPMs. The analysis reveals a diverse set of open problems. The main issue concerns the quality and number of learned prototypes. Further challenges include limited generalization across tasks and contexts, as well as methodological shortcomings such as non-standardized evaluation. Five broad research directions are identified: improving predictive performance, developing theoretically grounded architectures, establishing frameworks for human-AI collaboration, aligning models with human concepts, and defining robust metrics and benchmarks for evaluation. The survey aims to stimulate furthe
We present ProtoConcepts, a method for interpretable image classification combining deep learning and case-based reasoning using prototypical parts. Existing work in prototype-based image classification uses a ``this looks like that'' reasoning process, which dissects a test image by finding prototypical parts and combining evidence from these prototypes to make a final classification. However, all of the existing prototypical part-based image classifiers provide only one-to-one comparisons, where a single training image patch serves as a prototype to compare with a part of our test image. With these single-image comparisons, it can often be difficult to identify the underlying concept being compared (e.g., ``is it comparing the color or the shape?''). Our proposed method modifies the architecture of prototype-based networks to instead learn prototypical concepts which are visualized using multiple image patches. Having multiple visualizations of the same prototype allows us to more easily identify the concept captured by that prototype (e.g., ``the test image and the related training patches are all the same shade of blue''), and allows our model to create richer, more interpretab
When we are faced with challenging image classification tasks, we often explain our reasoning by dissecting the image, and pointing out prototypical aspects of one class or another. The mounting evidence for each of the classes helps us make our final decision. In this work, we introduce a deep network architecture -- prototypical part network (ProtoPNet), that reasons in a similar way: the network dissects the image by finding prototypical parts, and combines evidence from the prototypes to make a final classification. The model thus reasons in a way that is qualitatively similar to the way ornithologists, physicians, and others would explain to people on how to solve challenging image classification tasks. The network uses only image-level labels for training without any annotations for parts of images. We demonstrate our method on the CUB-200-2011 dataset and the Stanford Cars dataset. Our experiments show that ProtoPNet can achieve comparable accuracy with its analogous non-interpretable counterpart, and when several ProtoPNets are combined into a larger network, it can achieve an accuracy that is on par with some of the best-performing deep models. Moreover, ProtoPNet provides a level of interpretability that is absent in other interpretable deep models.
The results of spectral observations of NGC 3516 with the 2-m telescope of the Shamakhy Astrophysical Observatory during 2016-2020 are presented. In the first half of 2016, the intensive broad component Hbeta was found, which indicates a spectral type change compared to 2014, when the broad component was almost invisible. In the second half of 2016, the broad component H$β$ again was weakened and was practically not observed, remaining as weak until the end of 2019. At the end of 2019, the broad component Hbeta strengthened again, and in May 2020 reached a typical level for the high state of the object. During 2016-2020 we observed several changing looks of NGC 3516.
Perceptual judgment of image similarity by humans relies on rich internal representations ranging from low-level features to high-level concepts, scene properties and even cultural associations. However, existing methods and datasets attempting to explain perceived similarity use stimuli which arguably do not cover the full breadth of factors that affect human similarity judgments, even those geared toward this goal. We introduce a new dataset dubbed Totally-Looks-Like (TLL) after a popular entertainment website, which contains images paired by humans as being visually similar. The dataset contains 6016 image-pairs from the wild, shedding light upon a rich and diverse set of criteria employed by human beings. We conduct experiments to try to reproduce the pairings via features extracted from state-of-the-art deep convolutional neural networks, as well as additional human experiments to verify the consistency of the collected data. Though we create conditions to artificially make the matching task increasingly easier, we show that machine-extracted representations perform very poorly in terms of reproducing the matching selected by humans. We discuss and analyze these results, sugge
In 1981, David Mermin described a cleverly simplified version of Bell's theorem. It pointed out in a straightforward way that interpreting entanglement from a local realist point of view can be problematic. I propose here an extended version of Mermin's device that can actually be given a simple local realist interpretation through a sample selection bias, and I argue that we still have no scientific reason to believe that the moon could possibly not be there when nobody looks.
Modeling emotional-cognition is in a nascent stage and therefore wide-open for new ideas and discussions. In this paper the author looks at the modeling problem by bringing in ideas from axiomatic mathematics, information theory, computer science, molecular biology, non-linear dynamical systems and quantum computing and explains how ideas from these disciplines may have applications in modeling emotional-cognition.
We introduce a quantum mechanical model of time travel which includes two figurative beam splitters in order to induce feedback to earlier times. This leads to a unique solution to the paradox where one could kill one's grandfather in that once the future has unfolded, it cannot change the past, and so the past becomes deterministic. On the other hand, looking forwards towards the future is completely probabilistic. This resolves the classical paradox in a philosophically satisfying manner.
We investigate the structure of the Marvel Universe collaboration network, where two Marvel characters are considered linked if they jointly appear in the same Marvel comic book. We show that this network is clearly not a random network, and that it has most, but not all, characteristics of "real-life" collaboration networks, such as movie actors or scientific collaboration networks. The study of this artificial universe that tries to look like a real one, helps to understand that there are underlying principles that make real-life networks have definite characteristics.
Multi-look acquisition is a widely used strategy for reducing speckle noise in coherent imaging systems such as digital holography. By acquiring multiple measurements, speckle can be suppressed through averaging or joint reconstruction, typically under the assumption that speckle realizations across looks are statistically independent. In practice, however, hardware constraints limit measurement diversity, leading to inter-look correlation that degrades the performance of conventional methods. In this work, we study the reconstruction of speckle-free reflectivity from complex-valued multi-look measurements in the presence of correlated speckle. We model the inter-look dependence using a first-order Markov process and derive the corresponding likelihood under a first-order Markov approximation, resulting in a constrained maximum likelihood estimation problem. To solve this problem, we develop an efficient projected gradient descent framework that combines gradient-based updates with implicit regularization via deep image priors, and leverages Monte Carlo approximation and matrix-free operators for scalable computation. Simulation results demonstrate that the proposed approach remain
Multiplicative Gamma noise is a signal-dependent degradation in coherent imaging; synthetic aperture radar (SAR) despeckling is its most prominent real-world instance. Existing diffusion denoisers parameterize their forward process by abstract signal-to-noise schedules rather than by the physical look number $L$, so different deployment scenarios typically require separately trained models, and transfer from synthetic Gamma training to real SAR remains challenging without clean ground truth. We introduce $γ$-Bridge, a look-parametric bridge whose schedule $L(t)$ connects the noisy observation at $L_{obs}$ to the clean limit through exact multiplicative Gamma marginals. Its closed-form Gamma--Lévy reverse posterior admits both stochastic and deterministic processes, while observation conditioning and a two-step consistency loss stabilize multi-step inference in the low-SNR single-look regime. Because bridge time directly represents $L$, one conditioned network can smart-start from any admissible input look and stop at a target look number. These two orthogonal controls enable zero-shot restoration over the full admissible grid after training only at $L_{obs} = 1$ on natural images w
Recent advances in computer vision have yielded models with strong performance on recognition benchmarks; however, significant gaps remain in comparison to human perception. One subtle ability is to judge whether an image looks like a given object without being an instance of that object. We study whether vision-language models such as CLIP capture this distinction. We curated a dataset named RoLA (Real or Lookalike) of real and lookalike exemplars (e.g., toys, statues, drawings, pareidolia) across multiple categories, and first evaluate a prompt-based baseline with paired "real"/"lookalike" prompts. We then estimate a direction in CLIP's embedding space that moves representations between real and lookalike. Applying this direction to image and text embeddings improves discrimination in cross-modal retrieval on Conceptual12M, and also enhances captions produced by a CLIP prefix captioner.
We study reinforcement learning (RL) with transition look-ahead, where the agent may observe which states would be visited upon playing any sequence of $\ell$ actions before deciding its course of action. While such predictive information can drastically improve the achievable performance, we show that using this information optimally comes at a potentially prohibitive computational cost. Specifically, we prove that optimal planning with one-step look-ahead ($\ell=1$) can be solved in polynomial time through a novel linear programming formulation. In contrast, for $\ell \geq 2$, the problem becomes NP-hard. Our results delineate a precise boundary between tractable and intractable cases for the problem of planning with transition look-ahead in reinforcement learning.
Progress in vision research has been slower downstream than upstream of primary visual cortex (V1). Traditional frameworks have largely overlooked a central constraint: only a tiny fraction of retinal input is recognized. Thus, to a first approximation, vision is better formulated as looking and seeing through a bottleneck. Looking, mainly by the peripheral visual field, selects visual information to enter this bottleneck, largely via gaze shifts that center selected contents at fovea. Seeing, mainly by the central visual field, recognizes this content. Converging evidence suggests that V1 initiates the bottleneck and contributes to looking by generating a bottom-up saliency map that guides saccades exogenously, and that top-down feedback along the visual pathway, targeting mainly the representation of the central visual field, refines seeing. Progress will accelerate through falsifiable theories that explicitly link behavior with neural substrates, and by experimental designs that avoid forced fixation and precisely track gaze.
Changing-look active galactic nuclei (AGNs) are a special class of AGNs that change their spectral type from type 1 to type 2 or vice versa. In recent years, a number of changing-look blazars (CLBs) were also reported, which transition between flat-spectrum radio quasars and BL Lacs. The physical properties of CLBs are still unclear. Using the $mclust$ R package for Gaussian Mixture Modeling, we performed a clustering analysis for a sample of 105 CLBs selected from the literature. Three kinds of analysis found that CLBs lie in between the parameter distributions of FSRQs and BL Lacs: (i) univariate analysis; (ii) bivariate analysis; and (iii) multivariate analysis, carried out with a dimension reduction approach of the physical properties of the three types of blazars. Our results suggest that CLBs belong to a transition type between FSRQs and BL Lacs, which may be regulated by the change of accretion process and may be similar to other changing-look AGNs.
This report presents our team's 'PCIE_LAM' solution for the Ego4D Looking At Me Challenge at CVPR2024. The main goal of the challenge is to accurately determine if a person in the scene is looking at the camera wearer, based on a video where the faces of social partners have been localized. Our proposed solution, InternLSTM, consists of an InternVL image encoder and a Bi-LSTM network. The InternVL extracts spatial features, while the Bi-LSTM extracts temporal features. However, this task is highly challenging due to the distance between the person in the scene and the camera movement, which results in significant blurring in the face image. To address the complexity of the task, we implemented a Gaze Smoothing filter to eliminate noise or spikes from the output. Our approach achieved the 1st position in the looking at me challenge with 0.81 mAP and 0.93 accuracy rate. Code is available at https://github.com/KanokphanL/Ego4D_LAM_InternLSTM
This paper proposes a look ahead text understanding problem with look ahead section identification (LASI) as an example. This problem may appear in generative AI as well as human interactions, where we want to understand the direction of a developing text or conversation. We tackle the problem using transformer-based LLMs. We show that LASI is more challenging than classic section identification (SI). We argue that both bidirectional contextual information (e.g., BERT) and unidirectional predictive ability (e.g., GPT) will benefit the task. We propose two approaches to stitch together BERT and GPT. Experiments show that our approach outperforms the established models, especially when there is noise in the text (which is often the case for developing text in generative AI). Our paper sheds light on other look ahead text understanding tasks that are important to social media, such as look ahead sentiment classification, and points out the opportunities to leverage pre-trained LLMs through stitching.