Pith. sign in

REVIEW 4 major objections 5 minor 39 references

Application of Multimodal Large Language Models in Autonomous Driving

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read A fine-tuned multimodal language model can act as an autonomous-driving decision agent.

desk verdict Thin extension of an existing CogVLM driving agent; the quantitative evidence contradicts the central claim and lacks any statistical support. read the letter →

arxiv 2412.16410 v2 pith:J4BCQJZP submitted 2024-12-21 cs.CL

classification cs.CL
keywords autonomousdrivingmultimodallargelanguagemodelchain-of-thoughtvisualquestionansweringfine-tuningCogVLM2highway-envsceneunderstanding
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that a multimodal large language model, fine-tuned on just 100 visual question-answer examples drawn from BDD100K and KITTI, can serve as a practical autonomous-driving decision agent. The key move is to decompose driving decisions into three chain-of-thought stages—scene understanding, prediction, and decision—so the model produces interpretable reasoning before choosing an action. The paper reports that this fine-tuned CogVLM2 agent achieves near-zero failure rates and lower inefficiency than reinforcement learning and model-predictive-control baselines in highway-env simulations, while also handling corner cases with limited compute. A sympathetic reader would care because this points toward local, interpretable, few-shot deployment of language-model agents in vehicles rather than cloud-scale black boxes.

What carries the argument

The central machinery is the three-stage chain-of-thought that splits driving decisions into scene understanding, prediction, and decision, paired with a 100-image Visual Question Answering dataset built from BDD100K and KITTI to fine-tune CogVLM2. The CoT is what produces the step-by-step interpretable output that the paper credits with reducing model illusions and improving attention; the VQA dataset is what adapts a general-purpose MLLM to the driving domain. The action decoder reads the final stage of the answer to translate the model's textual output into driving actions.

What would settle it

Build a strict evaluation where the 100 VQA samples are split into train and test sets, hold out the demo images, and measure scene-understanding accuracy and simulation failure probability; if performance collapses on held-out images from the same domains, the claim of generalizable driving competence is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that an MLLM-driven framework effectively executes AD tasks with limited computer resources, few shots, multi-modality, and complex scenarios. Specifically, a CogVLM2 model fine-tuned on a hand-annotated and LLM-refined VQA dataset, guided by a three-stage chain-of-thought, can understand road scenes, predict other agents' behavior, and choose driving actions. The evidence is qualitative demonstrations on real road photos plus quantitative comparisons in three highway-env scenarios (intersection, roundabout, highway), where the proposed agent shows lower failure likelihood than RL, MPC, and the earlier [10] CogVLM agent in most settings.

Load-bearing premise

The load-bearing premise is that the qualitative demonstrations in Figures 4, 6, and 7 generalize beyond the specific 100-image training set, since the paper reports no train/test split and the demo images come from the same source datasets as fine-tuning.

Editorial extensions

If this is right

  • If the central claim holds, a small fine-tuned MLLM with a three-stage CoT can be deployed on an onboard GPU-like workstation rather than only in the cloud.
  • The three-stage decomposition makes decisions auditable: a driver or inspector can read the model's scene understanding, prediction, and final choice step by step.
  • In the reported simulations, the MLLM agent achieves a 0% failure rate in intersection and roundabout scenarios, compared with 4–10% for the RL and MPC baselines.
  • The paper positions this as an initial step toward safe, few-shot, local, and interpretable MLLM-driven AV systems.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Scaling the VQA dataset beyond 100 images would test whether the reported gains come from fine-tuning itself or from the CoT structure; the paper leaves this open.
  • The paper's demonstrations are still-image only, so a natural extension is to test whether the same CoT decomposition transfers to video-based or LiDAR-fused scene understanding.
  • Because the VQA dataset is English-only, the framework's claims are tied to English-language instruction following, a point the paper acknowledges in its limitations.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes fine-tuning the CogVLM2 multimodal large language model on a 100-image VQA dataset for autonomous driving, using a three-stage chain-of-thought (scene understanding, prediction, decision-making), and evaluates the resulting agent in Highway-env against RL, MPC, and the earlier CogVLM agent of reference [10]. The central claim is that this MLLM-based framework effectively executes AD tasks with limited resources and few-shot capability, and that MLLM-driven AD systems provide significant performance improvements. The evidence consists of qualitative examples for scene understanding, prediction, and decision-making, plus the simulation metrics in Table I.

Significance. If the central claim were established, a fine-tuned open VLM with chain-of-thought prompting could be a practically relevant interpretable AD decision agent. The construction of an AD-specific VQA dataset and the decomposition of decisions into three CoT stages are sensible ideas. However, the paper provides no held-out evaluation, no quantitative metrics for the understanding/prediction/decision tasks, no statistical support for the simulation results, and no code or data release. The simulation table is internally inconsistent and, on the highway scenario, the proposed method is worse than the prior method on two of three metrics. As presented, the evidence does not support the claimed significant improvements, and the contribution is not yet verifiable.

major comments (4)
  1. [Table I, Section IV-F] The highway row of Table I directly contradicts the paper's central claim of "significant performance improvements": for Highway, the proposed method reports Fail 0%, Inefficiency 5%, Average Time 23.6s, while the prior CogVLM agent [10] reports 0%, 2%, and 22.3s. The proposed method is strictly worse on Inefficiency and Average Time and equal on Fail. The paper gives no error bars, no number of repetitions, and no statistical test, so the quantitative superiority claim is unsupported. This is load-bearing for the first and third contribution bullets.
  2. [Table I, Section IV-F] The reported metric values are not consistent with the stated experimental setup. The paper says 50 traffic scenarios were tested, yet the roundabout row reports Inefficiency 2.3% and the merge row reports 4.7%, which cannot arise from integer counts of ineffective actions out of 50 trials unless the metric is defined differently than stated. The authors must define each metric precisely, report raw counts and denominators, and clarify how "Inefficiency" and "Average Time" are computed.
  3. [Sections IV-C, IV-D, IV-E, Figures 4, 6, 7] The scene understanding, prediction, and decision-making results are supported only by qualitative example images. The paper never reports a train/test split for the 100-image VQA dataset, which was built with ChatGPT-4 and Gemini annotations and used to fine-tune the model. The demonstration images in Figures 4, 6, and 7 appear to come from the same BDD100K/KITTI source domains as the fine-tuning data. Without a held-out split, the qualitative successes may reflect memorization rather than driving competence. A quantitative evaluation on a disjoint test set, with per-task accuracy or agreement metrics, is needed.
  4. [Section III-A] The action decoder is mentioned only in Section III-A as a mechanism for converting the model's third-stage answer into simulator actions, but no implementation details, action-space definition, or validation are provided. Without a reproducible mapping from language outputs to Highway-env actions, the simulation results in Table I cannot be independently verified or compared with the prior CogVLM agent.
minor comments (5)
  1. [Abstract, Section IV-E] The term "Virtual Question Answering" in the abstract and Section III-A should be "Visual Question Answering," and the word "wippers" in Section IV-E should be "wipers."
  2. [Section IV-E versus Section III-A] Section IV-E refers to "the thought process outlined in Sec III-B," but the chain-of-thought construction is described in Section III-A; the cross-reference is inconsistent.
  3. [Section II-C] There is a citation formatting error: "GPT-4 [ [31]" should be "GPT-4 [31]."
  4. [Section IV-F] The Highway-env environment is cited as reference [23] in Section IV-F but as reference [14] in the introduction; the numbering should be made consistent.
  5. [Section IV-A] The hardware description lists eight NVIDIA A100 GPUs but no training details such as fine-tuning epochs, learning rate, batch size, or use of parameter-efficient methods; this makes the fine-tuning procedure difficult to reproduce or assess.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: the paper fine-tunes CogVLM2 on a self-built VQA set and evaluates qualitatively and in simulation; no load-bearing claim is shown to reduce to its inputs by construction.

full rationale

The paper's chain is: build a VQA dataset using ChatGPT-4/Gemini annotations, fine-tune CogVLM2, demonstrate scene understanding/prediction/decision on sample images, and run highway-env simulations. I checked each link for definitional circularity, fitted-input-as-prediction, and self-citation load-bearing patterns. The VQA fine-tuning is an empirical training procedure, not a derivation; the qualitative demonstrations do not state that they are on held-out images, but the paper also does not show that the demonstrated outputs are the training targets, so the memorization concern is a validity risk rather than a constructional identity. The simulation uses an unstated action decoder, which is a reproducibility gap, not a circular step. Self-citations [37] and [39] are peripheral language-limitation remarks and are not load-bearing. The comparison to [10] shows the proposed method is worse on the highway scenario, but an adverse result is not circularity. Therefore no step can be exhibited where a prediction equals its input by construction, so the circularity score is 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a 100-example dataset, the action decoding protocol, and the representativeness of the qualitative and simulated scenarios. All of these are domain assumptions specific to this paper; none is backed by independent evidence, and the dataset/code are not released.

free parameters (4)
  • dataset_size = 100 images
    Chosen by hand; no analysis of how dataset size affects fine-tuning; all qualitative evaluation uses this pool.
  • annotation_iterations = 4
    Number of annotation/review cycles set by the author; no evidence it is sufficient.
  • chain_of_thought_stages = 3 (scene understanding, prediction, decision)
    Fixed decomposition used in prompts and evaluation; no ablation or justification given beyond a single image comparison.
  • simulation_scenarios = 50 per environment
    Number of test scenarios chosen by hand; no statistical power analysis or error bars.
assumptions (4)
  • domain assumption The 100-image VQA dataset built from BDD100K and KITTI with ChatGPT-4/Gemini annotations is representative of real-world AD corner cases.
    Used as the sole fine-tuning and demonstration data; no independent validation set, Section III-A.
  • domain assumption Textual decisions from the MLLM can be decoded into executable driving actions in highway-env.
    Simulation results depend on an action decoder whose implementation is not described, Section IV-E/F.
  • domain assumption Qualitative examples in Figs. 4, 6, and 7 demonstrate generalization rather than training-set recall.
    No train/test split is reported, Section IV-C/D/E.
  • domain assumption CogVLM2's better performance on general VLM benchmarks transfers to autonomous driving.
    Section IV-B motivates the model choice with TextVQA/DocVQA results, not AD-specific measurements.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Application of Multimodal Large Language Models in Autonomous Driving." pith.science (2026). https://pith.science/paper/J4BCQJZP

@misc{pith2026241216410,
  author       = {Pith},
  title        = {Pith review of: Application of Multimodal Large Language Models in Autonomous Driving},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J4BCQJZP}},
  note         = {Machine review of arXiv:2412.16410}
}
read the original abstract

In this era of technological advancements, several cutting-edge techniques are being implemented to enhance Autonomous Driving (AD) systems, focusing on improving safety, efficiency, and adaptability in complex driving environments. However, AD still faces some problems including performance limitations. To address this problem, we conducted an in-depth study on implementing the Multi-modal Large Language Model. We constructed a Virtual Question Answering (VQA) dataset to fine-tune the model and address problems with the poor performance of MLLM on AD. We then break down the AD decision-making process by scene understanding, prediction, and decision-making. Chain of Thought has been used to make the decision more perfectly. Our experiments and detailed analysis of Autonomous Driving give an idea of how important MLLM is for AD.

Figures

Figures reproduced from arXiv: 2412.16410 by the authors.

Figure 1
Figure 1. (a) Road surfaces with water, which may result in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The model’s step-by-step thinking chain generates information in a progressive manner, leading to more interpretable [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Visual Question Answer (VQA) dataset appropriately rating traffic conditions, and correctly detecting cars and people in the photos. D. Prediction based on Multi-modal LLM We tested our model’s predictive skills by selecting scenar￾ios with branched roads and decreased eyesight owing to rain, as shown in [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Used the fine-tuned CogVLM2 to describe traffic environments. The [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The configuration of NVIDIA GPU workstation [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Prediction of traffic environments using fine-tuned CogVLM2. Correctly understood responses are indicated by purple [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: The left image is showing decision with Chain of Thought(CoT) and the right one is without CoT [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: We compared our technique against RL and MPC in three different scenarios: roundabout, intersection, and highway. [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: Road merging identify key elements in the environment, such as: Road Types: Dual and single carriageways. Traffic Conditions: Real-time assessment of congestion or traffic flow. Objects: Accurate recognition of vehicles, pedestrians, and other road users. Contextual De…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 16 canonical work pages

  1. [10]

    An advanced driving agent with the multimodal large language model for autonomous vehicles,

    J. Chen and S. Lu, “An advanced driving agent with the multimodal large language model for autonomous vehicles,” in 2024 IEEE International Conference on Mobility, Operations, Services and Technologies (MOST). IEEE, 2024, pp. 1–11

  2. [1]

    A survey on multimodal large language models,

    S. Yin, C. Fu, S. Zhao, K. Li, X. Sun, T. Xu, and E. Chen, “A survey on multimodal large language models,” National Science Review , p. nwae403, 2024

  3. [2]

    Tesla autopilot: semi autonomous driving, an uptick for future autonomy,

    S. Ingle and M. Phute, “Tesla autopilot: semi autonomous driving, an uptick for future autonomy,” International Research Journal of Engineering and Technology , vol. 3, no. 9, pp. 369–372, 2016

  4. [3]

    Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset,

    S. Ettinger, S. Cheng, B. Caine, C. Liu, H. Zhao, S. Pradhan, Y . Chai, B. Sapp, C. R. Qi, Y . Zhou et al. , “Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 9710–9719

  5. [4]

    Application of baidu apollo open platform in a course of control simulation experiments,

    M. Feng and H. Zhang, “Application of baidu apollo open platform in a course of control simulation experiments,” Computer Applications in Engineering Education , vol. 30, no. 3, pp. 892–906, 2022

  6. [5]

    Dynamically conservative self-driving planner for long-tail cases,

    W. Zhou, Z. Cao, N. Deng, X. Liu, K. Jiang, and D. Yang, “Dynamically conservative self-driving planner for long-tail cases,” IEEE Transactions on Intelligent Transportation Systems , vol. 24, no. 3, pp. 3476–3488, 2022

  7. [6]

    Autonomous driving with deep learning: A survey of state-of-art technologies. arxiv 2020,

    Y . Huang and Y . Chen, “Autonomous driving with deep learning: A survey of state-of-art technologies. arxiv 2020,” arXiv preprint arXiv:2006.06091, 2006

  8. [7]

    Planning-oriented autonomous driving,

    Y . Hu, J. Yang, L. Chen, K. Li, C. Sima, X. Zhu, S. Chai, S. Du, T. Lin, W. Wang et al. , “Planning-oriented autonomous driving,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17 853–17 862

Show all 39 references
  1. [8]

    Milestones in autonomous driving and intelligent vehicles: Survey of surveys,

    L. Chen, Y . Li, C. Huang, B. Li, Y . Xing, D. Tian, L. Li, Z. Hu, X. Na, Z. Li et al., “Milestones in autonomous driving and intelligent vehicles: Survey of surveys,” IEEE Transactions on Intelligent V ehicles , vol. 8, no. 2, pp. 1046–1056, 2022

  2. [9]

    Is chatgpt a good translator? yes with gpt-4 as the engine,

    W. Jiao, W. Wang, J.-t. Huang, X. Wang, S. Shi, and Z. Tu, “Is chatgpt a good translator? yes with gpt-4 as the engine,” arXiv preprint arXiv:2301.08745, 2023

  3. [11]

    Baichuan 2: Open large-scale language models,

    A. Yang, B. Xiao, B. Wang, B. Zhang, C. Bian, C. Yin, C. Lv, D. Pan, D. Wang, D. Yanet al., “Baichuan 2: Open large-scale language models,” arXiv preprint arXiv:2309.10305 , 2023

  4. [12]

    Complexity- based prompting for multi-step reasoning,

    Y . Fu, H. Peng, A. Sabharwal, P. Clark, and T. Khot, “Complexity- based prompting for multi-step reasoning,” in The Eleventh International Conference on Learning Representations , 2022

  5. [13]

    Using multimodal large language models (mllms) for automated detection of traffic safety-critical events,

    M. Abu Tami, H. I. Ashqar, M. Elhenawy, S. Glaser, and A. Rakotoni- rainy, “Using multimodal large language models (mllms) for automated detection of traffic safety-critical events,” V ehicles, vol. 6, no. 3, pp. 1571–1590, 2024

  6. [14]

    An environment for autonomous driving decision-making,

    E. Leurent, “An environment for autonomous driving decision-making,” https://github.com/eleurent/highway-env, 2018

  7. [15]

    Taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles,

    S. International, “Taxonomy and definitions for terms related to driving automation systems for on-road motor vehicles,” https://www.sae.org/ standards/content/j3016 202104/, 2021, accessed: December 7, 2024

  8. [16]

    End-to-end driving via conditional imitation learning,

    F. Codevilla, M. M ¨uller, A. L ´opez, V . Koltun, and A. Dosovitskiy, “End-to-end driving via conditional imitation learning,” in 2018 IEEE international conference on robotics and automation (ICRA) . IEEE, 2018, pp. 4693–4700

  9. [17]

    A review on safety failures, security attacks, and available countermeasures for autonomous vehicles,

    J. Cui, L. S. Liew, G. Sabaliauskaite, and F. Zhou, “A review on safety failures, security attacks, and available countermeasures for autonomous vehicles,” Ad Hoc Netw. , vol. 90, no. C, Jul. 2019. [Online]. Available: https://doi.org/10.1016/j.adhoc.2018.12.006

  10. [18]

    A survey of autonomous driving: Common practices and emerging technologies,

    E. Yurtsever, J. Lambert, A. Carballo, and K. Takeda, “A survey of autonomous driving: Common practices and emerging technologies,” IEEE access , vol. 8, pp. 58 443–58 469, 2020

  11. [19]

    A formal basis for the heuristic determination of minimum cost paths,

    P. E. Hart, N. J. Nilsson, and B. Raphael, “A formal basis for the heuristic determination of minimum cost paths,” IEEE transactions on Systems Science and Cybernetics , vol. 4, no. 2, pp. 100–107, 1968

  12. [20]

    Robot motion planning: A distributed representation approach,

    J. Barraquand and J.-C. Latombe, “Robot motion planning: A distributed representation approach,” The International Journal of Robotics Re- search, vol. 10, no. 6, pp. 628–649, 1991

  13. [21]

    Randomized kinodynamic planning,

    S. M. LaValle and J. J. Kuffner Jr, “Randomized kinodynamic planning,” The international journal of robotics research , vol. 20, no. 5, pp. 378– 400, 2001

  14. [22]

    Sampling-based algorithms for optimal motion planning,

    S. Karaman and E. Frazzoli, “Sampling-based algorithms for optimal motion planning,” The international journal of robotics research, vol. 30, no. 7, pp. 846–894, 2011

  15. [23]

    Motion planning for autonomous driving: The state of the art and future perspectives,

    S. Teng, X. Hu, P. Deng, B. Li, Y . Li, Y . Ai, D. Yang, L. Li, Z. Xuanyuan, F. Zhu et al., “Motion planning for autonomous driving: The state of the art and future perspectives,” IEEE Transactions on Intelligent V ehicles , vol. 8, no. 6, pp. 3692–3711, 2023

  16. [24]

    Multi-task conditional imitation learning for autonomous navigation at crowded intersections,

    Z. Zhu and H. Zhao, “Multi-task conditional imitation learning for autonomous navigation at crowded intersections,” arXiv preprint arXiv:2202.10124, 2022

  17. [25]

    Cola-hrl: Continuous-lattice hierarchical reinforcement learning for autonomous driving,

    L. Gao, Z. Gu, C. Qiu, L. Lei, S. E. Li, S. Zheng, W. Jing, and J. Chen, “Cola-hrl: Continuous-lattice hierarchical reinforcement learning for autonomous driving,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) . IEEE, 2022, pp. 13 143–13 150

  18. [26]

    Parallel motion planning: Learning a deep planning model against emergencies,

    L. Chen, X. Hu, B. Tang, and D. Cao, “Parallel motion planning: Learning a deep planning model against emergencies,” IEEE Intelligent Transportation Systems Magazine , vol. 11, no. 1, pp. 36–41, 2018

  19. [27]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,” Advances in Neural Information Processing Systems, 2017

  20. [28]

    Bert: Pre-training of deep bidirectional transformers for language understanding,

    J. Devlin, “Bert: Pre-training of deep bidirectional transformers for language understanding,” arXiv preprint arXiv:1810.04805 , 2018

  21. [29]

    Scaling laws for neural language models,

    J. Kaplan, S. McCandlish, T. Henighan, T. B. Brown, B. Chess, R. Child, S. Gray, A. Radford, J. Wu, and D. Amodei, “Scaling laws for neural language models,” arXiv preprint arXiv:2001.08361 , 2020

  22. [30]

    Language models are few-shot learners advances in neural information processing systems 33,

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askellet al., “Language models are few-shot learners advances in neural information processing systems 33,” 2020

  23. [31]

    Gpt-4 technical report,

    J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F. L. Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkat et al. , “Gpt-4 technical report,” arXiv preprint arXiv:2303.08774 , 2023

  24. [32]

    Llama 2: Open foundation and fine-tuned chat models,

    H. Touvron, L. Martin, K. Stone, P. Albert, A. Almahairi, Y . Babaei, N. Bashlykov, S. Batra, P. Bhargava, S. Bhosale et al. , “Llama 2: Open foundation and fine-tuned chat models,” arXiv preprint arXiv:2307.09288, 2023

  25. [33]

    Qwen technical report,

    J. Bai, S. Bai, Y . Chu, Z. Cui, K. Dang, X. Deng, Y . Fan, W. Ge, Y . Han, F. Huang et al. , “Qwen technical report,” arXiv preprint arXiv:2309.16609, 2023

  26. [34]

    Textbooks are all you need ii: phi-1.5 technical report,

    Y . Li, S. Bubeck, R. Eldan, A. Del Giorno, S. Gunasekar, and Y . T. Lee, “Textbooks are all you need ii: phi-1.5 technical report,” arXiv preprint arXiv:2309.05463, 2023

  27. [35]

    Bdd100k: A diverse driving dataset for heterogeneous multitask learning,

    F. Yu, H. Chen, X. Wang, W. Xian, Y . Chen, F. Liu, V . Madhavan, and T. Darrell, “Bdd100k: A diverse driving dataset for heterogeneous multitask learning,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2020, pp. 2636–2645

  28. [36]

    Are we ready for autonomous driving? the kitti vision benchmark suite,

    A. Geiger, P. Lenz, and R. Urtasun, “Are we ready for autonomous driving? the kitti vision benchmark suite,” in Conference on Computer Vision and Pattern Recognition (CVPR) , 2012

  29. [37]

    Sentiment polarity analysis of bangla food reviews using machine and deep learning algorithms,

    A. Amin, A. Sarkar, M. M. Islam, A. A. Miazee, M. R. Islam, and M. M. Hoque, “Sentiment polarity analysis of bangla food reviews using machine and deep learning algorithms,” in 2024 3rd International Conference on Advancement in Electrical and Electronic Engineering (ICAEEE), ...

  30. [38]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Loet al., “Segment anything,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 4015–4026

  31. [39]

    Enhancing bangla language next word prediction and sentence completion through extended rnn with bi-lstm model on n-gram language,

    M. R. Islam, A. Amin, and A. N. Zereen, “Enhancing bangla language next word prediction and sentence completion through extended rnn with bi-lstm model on n-gram language,” in 2024 3rd International Conference on Advancement in Electrical and Electronic Engineering (ICAEEE), 2...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.