Nikita Odinets
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Nikita Odinets has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Anastasia Sarycheva
Institute for Information Transmission Problems of the Russian Academy of Sciences, 127051 Moscow, Russia
3 shared publications - Dmitry Sidorchuk
Institute for Information Transmission Problems, the Russian Academy of Sciences, 127051 Moscow, Russia
3 shared publications - Artem Nikonorov
Samara National Research University, 34 Moskovskoe shosse, Samara 443086, Russia
2 shared publications - Dmitrii P. Nikolaev
Federal Research Center “Computer Science and Control” of the Russian Academy of Sciences, Moscow 119333, Russia
2 shared publications - Ekaterina Dudenko
Federal Research Center Computer Science and Control, the Russian Academy of Sciences, 119333 Moscow, Russia
2 shared publications - Mikhail Zhidkov
Institute for Information Transmission Problems, the Russian Academy of Sciences, 127051 Moscow, Russia
2 shared publications
Latest Publications
- Open Access
- 22 June 2026
- Published by MDPI AG in Preprints.org
BMRS: Bongard–Maximov Problems for Remote Sensing
Vision-language models (VLMs) are increasingly used in remote sensing (RS), yet their ability to perform abstract visual reasoning remains poorly understood. Existing RS benchmarks primarily evaluate perception-oriented capabilities, such as scene classification, object detection, image captioning, and visual question answering, providing limited insight into higher-level reasoning. To address this gap, we introduce BMRS, the first remote-sensing benchmark based on the Bongard problem paradigm. The benchmark comprises 122 problems constructed from satellite and aerial imagery and organized into seven categories: Shape, Semantic, Presence, Spatial, Size, Number, and Same. Unlike conventional perception benchmarks, BMRS requires models to infer abstract concepts from multiple images, thereby evaluating few-shot learning, concept induction, analogy making, and relational reasoning. To establish a human baseline, we conducted a study involving 113 participants, who achieved an average accuracy of 74.5%. We then evaluated three groups of models: proprietary large-scale VLMs, open-source general-purpose VLMs, and remote sensing vision-language models (RSVLMs). The strongest proprietary models, ChatGPT and Gemini, achieved accuracies of 89.3% and 86.9%, respectively, surpassing human performance. In contrast, the best open-source general-purpose model achieved 42.6% accuracy, while the strongest RSVLM reached only 20.5%. Analysis across problem categories revealed that Semantic problems were the easiest for both humans and models, indicating that object-level semantic understanding transfers effectively to the remote-sensing domain: several RSVLMs achieved performance comparable to substantially larger general-purpose models. Conversely, Spatial problems proved the most challenging, highlighting spatial-relational reasoning as a key limitation of current RSVLMs. BMRS provides a challenging benchmark for measuring progress toward reasoning-capable RSVLMs and offers new insights into the strengths and limitations of current model adaptation strategies.
- 22 June 2026
- Published by MDPI AG in Preprints.org
Open AccessBMRS: Bongard–Maximov Problems for Remote Sensing
Vision-language models (VLMs) are increasingly used in remote sensing (RS), yet their ability to perform abstract visual reasoning remains poorly understood. Existing RS benchmarks primarily evaluate perception-oriented capabilities, such as scene classification, object detection, image captioning, and visual question answering, providing limited insight into higher-level reasoning. To address this gap, we introduce BMRS, the first remote-sensing benchmark based on the Bongard problem paradigm. The benchmark comprises 122 problems constructed from satellite and aerial imagery and organized into seven categories: Shape, Semantic, Presence, Spatial, Size, Number, and Same. Unlike conventional perception benchmarks, BMRS requires models to infer abstract concepts from multiple images, thereby evaluating few-shot learning, concept induction, analogy making, and relational reasoning. To establish a human baseline, we conducted a study involving 113 participants, who achieved an average accuracy of 74.5%. We then evaluated three groups of models: proprietary large-scale VLMs, open-source general-purpose VLMs, and remote sensing vision-language models (RSVLMs). The strongest proprietary models, ChatGPT and Gemini, achieved accuracies of 89.3% and 86.9%, respectively, surpassing human performance. In contrast, the best open-source general-purpose model achieved 42.6% accuracy, while the strongest RSVLM reached only 20.5%. Analysis across problem categories revealed that Semantic problems were the easiest for both humans and models, indicating that object-level semantic understanding transfers effectively to the remote-sensing domain: several RSVLMs achieved performance comparable to substantially larger general-purpose models. Conversely, Spatial problems proved the most challenging, highlighting spatial-relational reasoning as a key limitation of current RSVLMs. BMRS provides a challenging benchmark for measuring progress toward reasoning-capable RSVLMs and offers new insights into the strengths and limitations of current model adaptation strategies.
- Open Access
- 26 November 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (12) , 422
HyperHazeOff: Hyperspectral Remote Sensing Image Dehazing Benchmark
Hyperspectral remote sensing images (HSIs) provide invaluable information for environmental and agricultural monitoring, yet they are often degraded by atmospheric haze, which distorts spatial and spectral content and hinders downstream analysis. Progress in hyperspectral dehazing has been limited by the absence of paired real-haze benchmarks; most prior studies rely on synthetic haze or unpaired data, restricting fair evaluation and generalization. We present HyperHazeOff, the first comprehensive benchmark for hyperspectral dehazing that unifies data, tasks, and evaluation protocols. It comprises (i) RRealHyperPDID, 110 scenes with paired real-haze and haze-free HSIs (plus RGB images), and (ii) RSyntHyperPDID, 2616 paired samples generated using a physically grounded haze formation model. The benchmark also provides agricultural field delineation and land classification annotations for downstream task quality assessment, standardized train/validation/test splits, preprocessing pipelines, baseline implementations, pretrained weights, and evaluation tools. Across six state-of-the-art methods (three RGB-based and three HSI-specific), we find that hyperspectral models trained on the widely used HyperDehazing dataset fail to generalize to real haze, while training on RSyntHyperPDID enables significant real-haze restoration by AACNet. HyperHazeOff establishes reproducible baselines and is openly available to advance research in hyperspectral dehazing.
- 26 November 2025
- Published by MDPI AG in Journal of Imaging
- Vol. 11 (12) , 422
Open AccessHyperHazeOff: Hyperspectral Remote Sensing Image Dehazing Benchmark
Hyperspectral remote sensing images (HSIs) provide invaluable information for environmental and agricultural monitoring, yet they are often degraded by atmospheric haze, which distorts spatial and spectral content and hinders downstream analysis. Progress in hyperspectral dehazing has been limited by the absence of paired real-haze benchmarks; most prior studies rely on synthetic haze or unpaired data, restricting fair evaluation and generalization. We present HyperHazeOff, the first comprehensive benchmark for hyperspectral dehazing that unifies data, tasks, and evaluation protocols. It comprises (i) RRealHyperPDID, 110 scenes with paired real-haze and haze-free HSIs (plus RGB images), and (ii) RSyntHyperPDID, 2616 paired samples generated using a physically grounded haze formation model. The benchmark also provides agricultural field delineation and land classification annotations for downstream task quality assessment, standardized train/validation/test splits, preprocessing pipelines, baseline implementations, pretrained weights, and evaluation tools. Across six state-of-the-art methods (three RGB-based and three HSI-specific), we find that hyperspectral models trained on the widely used HyperDehazing dataset fail to generalize to real haze, while training on RSyntHyperPDID enables significant real-haze restoration by AACNet. HyperHazeOff establishes reproducible baselines and is openly available to advance research in hyperspectral dehazing.
- Open Access
- 20 October 2025
- Published by MDPI AG in Preprints.org
HyperHazeOff: Hyperspectral Remote Sensing Image Dehazing Benchmark
Hyperspectral remote sensing imagery provides invaluable information for environmental and agricultural monitoring but remains highly susceptible to atmospheric haze. We introduce the first comprehensive benchmark for hyperspectral image dehazing featuring paired real-world hazy and haze-free remote sensing data. The Remote sensing Real-world Hyperspectral Paired Dehazing Image Dataset (RRealHyperPDID) includes 110 scenes with naturally occurring haze and corresponding clear references, enabling reliable evaluation of dehazing algorithms across both hyperspectral and RGB domains. To support model training and ablation studies, we present the Remote Sensing Synthetic Hyperspectral Paired Dehazing Image Dataset (RSyntHyperPDID), comprising 2,616 synthetically generated hazy-clear pairs. The benchmark also provides agricultural field delineation annotations, allowing assessment of dehazing effects on downstream remote sensing tasks. Extensive experiments with six state-of-the-art methods reveal fundamental limitations of existing hyperspectral dehazing approaches and highlight the need for realistic, domain-consistent data. Models trained on RSyntHyperPDID demonstrate better generalization to real hazy data, establishing a new baseline for hyperspectral dehazing.
- 20 October 2025
- Published by MDPI AG in Preprints.org
Open AccessHyperHazeOff: Hyperspectral Remote Sensing Image Dehazing Benchmark
Hyperspectral remote sensing imagery provides invaluable information for environmental and agricultural monitoring but remains highly susceptible to atmospheric haze. We introduce the first comprehensive benchmark for hyperspectral image dehazing featuring paired real-world hazy and haze-free remote sensing data. The Remote sensing Real-world Hyperspectral Paired Dehazing Image Dataset (RRealHyperPDID) includes 110 scenes with naturally occurring haze and corresponding clear references, enabling reliable evaluation of dehazing algorithms across both hyperspectral and RGB domains. To support model training and ablation studies, we present the Remote Sensing Synthetic Hyperspectral Paired Dehazing Image Dataset (RSyntHyperPDID), comprising 2,616 synthetically generated hazy-clear pairs. The benchmark also provides agricultural field delineation annotations, allowing assessment of dehazing effects on downstream remote sensing tasks. Extensive experiments with six state-of-the-art methods reveal fundamental limitations of existing hyperspectral dehazing approaches and highlight the need for realistic, domain-consistent data. Models trained on RSyntHyperPDID demonstrate better generalization to real hazy data, establishing a new baseline for hyperspectral dehazing.