1 code implementation • 19 Mar 2024 • Beomsu Kim, JaeMin Kim, Jeongsol Kim, Jong Chul Ye
Diffusion-based generative models excel in unconditional generation, as well as on applied tasks such as image editing and restoration.
no code implementations • 31 Jan 2024 • Geonung Kim, Beomsu Kim, Eunhyeok Park, Sunghyun Cho
As recent advancements in large-scale Text-to-Image (T2I) diffusion models have yielded remarkable high-quality image generation, diverse downstream Image-to-Image (I2I) applications have emerged.
1 code implementation • NeurIPS 2023 • Geon Yeong Park, Jeongsol Kim, Beomsu Kim, Sang Wan Lee, Jong Chul Ye
Despite the remarkable performance of text-to-image diffusion models in image generation tasks, recent studies have raised the issue that generated images sometimes cannot capture the intended semantic contents of the text prompts, which phenomenon is often called semantic misalignment.
1 code implementation • 24 May 2023 • Beomsu Kim, Gihyun Kwon, Kwanyoung Kim, Jong Chul Ye
Diffusion models are a powerful class of generative models which simulate stochastic differential equations (SDEs) to generate data from noise.
no code implementations • 20 May 2023 • Taekyung Kim, Jungwi Mun, Junwon Seo, Beomsu Kim, Seongil Hong
Active exploration, in which a robot directs itself to states that yield the highest information gain, is essential for efficient data collection and minimizing human supervision.
1 code implementation • 27 Jan 2023 • Sangyun Lee, Beomsu Kim, Jong Chul Ye
Based on the relationship between the forward process and the curvature, here we present an efficient method of training the forward process to minimize the curvature of generative trajectories without any ODE/SDE simulation.
1 code implementation • 11 Oct 2022 • Seungju Han, Beomsu Kim, Buru Chang
In this paper, we introduce a new automatic evaluation metric to measure the semantic diversity of generated responses.
1 code implementation • 29 Sep 2022 • Beomsu Kim, Jong Chul Ye
Diffusion models are powerful generative models that simulate the reverse of diffusion processes using score functions to synthesize data from noise.
no code implementations • 30 May 2022 • Beomsu Kim, Jong Chul Ye
Deep energy-based models (EBMs), which use deep neural networks (DNNs) as energy functions, are receiving increasing attention due to their ability to learn complex distributions.
1 code implementation • NAACL 2022 • Seungju Han, Beomsu Kim, Jin Yong Yoo, Seokjun Seo, SangBum Kim, Enkhbayar Erdenee, Buru Chang
To better reflect the style of the character, PDP builds the prompts in the form of dialog that includes the character's utterances as dialog history.
no code implementations • 21 Feb 2022 • Beomsu Kim, Junghoon Seo
Adversarial examples, crafted by adding imperceptible perturbations to natural inputs, can easily fool deep neural networks (DNNs).
1 code implementation • 10 Feb 2022 • Beomsu Kim, Jong Chul Ye
Contrastive learning is a method of learning visual representations by training Deep Neural Networks (DNNs) to increase the similarity between representations of positive pairs (transformations of the same image) and reduce the similarity between representations of negative pairs (transformations of different images).
1 code implementation • NLP4ConvAI (ACL) 2022 • Seungju Han, Beomsu Kim, Seokjun Seo, Enkhbayar Erdenee, Buru Chang
Extensive experiments demonstrate that our proposed training method alleviates the drawbacks of the existing exemplar-based generative models and significantly improves the performance in terms of appropriateness and informativeness.
1 code implementation • Findings (EMNLP) 2021 • Beomsu Kim, Seokjun Seo, Seungju Han, Enkhbayar Erdenee, Buru Chang
G2R consists of two distinct techniques of distillation: the data-level G2R augments the dialogue dataset with additional responses generated by the large-scale generative model, and the model-level G2R transfers the response quality score assessed by the generative model to the score of the retrieval model by the knowledge distillation loss.
2 code implementations • CVPR 2021 • Youngkyu Hong, Seungju Han, Kwanghee Choi, Seokjun Seo, Beomsu Kim, Buru Chang
Although this method surpasses state-of-the-art methods on benchmark datasets, it can be further improved by directly disentangling the source label distribution from the model prediction in the training phase.
Ranked #20 on Long-tail Learning on Places-LT
no code implementations • ECCV 2020 • Jonghwa Yim, Jisung Yoo, Won-joon Do, Beomsu Kim, Jihwan Choe
Unlike conventional style transfer, new technique FST can extract and transfer custom filter style from a filtered style image to a content image.
no code implementations • 19 Nov 2019 • Sungjoo Ha, Martin Kersner, Beomsu Kim, Seokjun Seo, Dongyoung Kim
When there is a mismatch between the target identity and the driver identity, face reenactment suffers severe degradation in the quality of the result, especially in a few-shot setting.
no code implementations • 4 Oct 2019 • Junghoon Seo, Seungwon Lee, Beomsu Kim, Taegyun Jeon
In this paper, we revisit the classical bootstrap aggregating approach in the context of modern transfer learning for data-efficient disaster damage detection.
3 code implementations • 8 Apr 2019 • Seungwoo Choi, Seokjun Seo, Beomjun Shin, Hyeongmin Byun, Martin Kersner, Beomsu Kim, Dongyoung Kim, Sungjoo Ha
In addition, we release the implementation of the proposed and the baseline models including an end-to-end pipeline for training models and evaluating them on mobile devices.
Ranked #14 on Keyword Spotting on Google Speech Commands (Google Speech Commands V2 12 metric)
1 code implementation • 27 Mar 2019 • Beomsu Kim, Junghoon Seo, Taegyun Jeon
Adversarial training is a training scheme designed to counter adversarial attacks by augmenting the training dataset with adversarial examples.
2 code implementations • 13 Feb 2019 • Beomsu Kim, Junghoon Seo, SeungHyun Jeon, Jamyoung Koo, Jeongyeol Choe, Taegyun Jeon
Saliency Map, the gradient of the score function with respect to the input, is the most basic technique for interpreting deep neural network decisions.
no code implementations • 27 Sep 2018 • Beomsu Kim, Junghoon Seo, Jeongyeol Choe, Jamyoung Koo, Seunghyeon Jeon, Taegyun Jeon
In this paper, we identify the cause of noisy saliency maps.
no code implementations • 8 Jun 2018 • Junghoon Seo, Jeongyeol Choe, Jamyoung Koo, Seunghyeon Jeon, Beomsu Kim, Taegyun Jeon
SmoothGrad and VarGrad are techniques that enhance the empirical quality of standard saliency maps by adding noise to input.
no code implementations • CVPR 2018 • Junhyug Noh, Soochan Lee, Beomsu Kim, Gunhee Kim
We propose methods of addressing two critical issues of pedestrian detection: (i) occlusion of target objects as false negative failure, and (ii) confusion with hard negative examples like vertical structures as false positive failure.