I am hesitating whether I should participate in this competition or not.
\nI have some experience in working with tabular data in kaggle competition. But I do not know much about how to handle non tabular data or deep learning(other than image recognition) so would it be a good competition for beginners for me? But I am very interested in this competition since I speak bengali.
I am hesitating whether I should participate in this competition or not.
\nI have some experience in working with tabular data in kaggle competition. But I do not know much about how to handle non tabular data or deep learning(other than image recognition) so would it be a good competition for beginners for me? But I am very interested in this competition since I speak bengali.
this is not for beginners, but I would highly suggest you participate and try to learn going through the discussions and code, if you have the time and interest, its just a different style of competition. best wishes!
", "rawMarkdown": "this is not for beginners, but I would highly suggest you participate and try to learn going through the discussions and code, if you have the time and interest, its just a different style of competition. best wishes!", "votes": null }, { "id": "2385922", "postDate": "08/11/2023 16:20:33", "content": "Hey!
\nThis competition lasts for 2 more months. You can set this as your challenge, and try working towards it. You can start off with available code and previous competition, look for existing models, and various methods used. Assuming you are consistent enough, you should learn a lot after this competition. You will pickup a lot of skills in using Large Models, Training with constraints, Basics of Language Models, How is deep learning used in speech and so forth. Remember to learn about the models you use, not just blindly use them.
\nHappy kaggling!
", "rawMarkdown": "Hey!\n\nThis competition lasts for 2 more months. You can set this as your challenge, and try working towards it. You can start off with available code and previous competition, look for existing models, and various methods used. Assuming you are consistent enough, you should learn a lot after this competition. You will pickup a lot of skills in using Large Models, Training with constraints, Basics of Language Models, How is deep learning used in speech and so forth. Remember to learn about the models you use, not just blindly use them.\n\nHappy kaggling!", "votes": null }, { "id": "2386707", "postDate": "08/12/2023 05:51:34", "content": "@shoabahamed I'm also a beginner, I joined this competition for learning purposes. I suggest you to join this competetion.
", "rawMarkdown": "shoabahamed I'm also a beginner, I joined this competition for learning purposes. I suggest you to join this competetion.", "votes": null }, { "id": "2386716", "postDate": "08/12/2023 05:56:26", "content": "@anivana I'm a beginner at this stage. I learn and apply Fourier transform and spectrogram. Any suggestion for model building? Which algorithm should I use?
", "rawMarkdown": "anivana I'm a beginner at this stage. I learn and apply Fourier transform and spectrogram. Any suggestion for model building? Which algorithm should I use?", "votes": null }, { "id": "2386847", "postDate": "08/12/2023 07:29:47", "content": "https://www.kaggle.com/competitions/bengaliai-speech/discussion/425250
\nStart with this, and you will learn what to do in short time :) best wishes.
", "rawMarkdown": "https://www.kaggle.com/competitions/bengaliai-speech/discussion/425250\n\nStart with this, and you will learn what to do in short time :) best wishes.", "votes": null }, { "id": "2387492", "postDate": "08/12/2023 17:40:19", "content": "As someone who perpetually participates and never wins but usually learns a lot, 100% yes. Worst case is you've clicked on a button to participate and get free access to data and a GPU on kaggle. The discussions will provide state of the art resources and usually links to courses.
\nI can't remember what competition it was, but I thought I saw a winner one time say they didn't even know what deep learning was when they started lol.
\nNever doubt yourself and as long as it's not too much of a time or mental struggle to affect your life, give it all you got and see where you end up.
", "rawMarkdown": "As someone who perpetually participates and never wins but usually learns a lot, 100% yes. Worst case is you've clicked on a button to participate and get free access to data and a GPU on kaggle. The discussions will provide state of the art resources and usually links to courses. \n\nI can't remember what competition it was, but I thought I saw a winner one time say they didn't even know what deep learning was when they started lol. \n\nNever doubt yourself and as long as it's not too much of a time or mental struggle to affect your life, give it all you got and see where you end up.", "votes": null }, { "id": "2390111", "postDate": "08/14/2023 11:33:33", "content": "I think what matters the most is your motivation. Is this competition excites you in terms of new skills and new ways of approaching problems? If yes, start right away. If you can keep your motivation until the last minute, you'll enjoy a lot as you proceed. And any struggles along the way won't feel like one but a new opportunity to move one step further. So. give it a try, explore the dataset, and decide.
\nIf you want to participate at the end of the day, here are two great resources to start tackling this problem in addition to the posts from the competition host(s) themselves:
\n\n\nhttps://www.youtube.com/playlist?list=PL-wATfeyAMNqIee7cH3q1bh4QJFAaeNv0
\n
\n \n\n
Especially the chapter 5 will give you all the necessary pieces to start with Automatic Speech Recognition
", "rawMarkdown": "I think what matters the most is your motivation. Is this competition excites you in terms of new skills and new ways of approaching problems? If yes, start right away. If you can keep your motivation until the last minute, you'll enjoy a lot as you proceed. And any struggles along the way won't feel like one but a new opportunity to move one step further. So. give it a try, explore the dataset, and decide.\n\nIf you want to participate at the end of the day, here are two great resources to start tackling this problem in addition to the posts from the competition host(s) themselves:\n\n- Here is a video series about Audio Signal Processing which can be definitely helpful for beginners in audio processing. It also includes code examples in Python.\n\n> https://www.youtube.com/playlist?list=PL-wATfeyAMNqIee7cH3q1bh4QJFAaeNv0\n\n- The presenter also has a website where he teaches about different applications of AI for audio data.\n\n> https://valeriovelardo.com/the-sound-of-ai/#\n\n- The video course from HuggingFace: https://huggingface.co/learn/audio-course/chapter0/introduction\n\nEspecially the chapter 5 will give you all the necessary pieces to start with Automatic Speech Recognition", "votes": null }, { "id": "2390318", "postDate": "08/14/2023 13:43:55", "content": "Thanks for the suggestions and links.
", "rawMarkdown": "Thanks for the suggestions and links.", "votes": null }, { "id": "2390368", "postDate": "08/14/2023 13:59:10", "content": "great answer and links, thanks!
", "rawMarkdown": "great answer and links, thanks!", "votes": null }, { "id": "2391941", "postDate": "08/15/2023 11:38:08", "content": "I am also a beginner when it comes to nontabular data and hoping to learn something new from this competition. This is the first time I'm working with any audio data and I have been trying to understand the basics of audio signals and how to process it since couple of days.
\nI have summarized all my learning in the following notebook: https://www.kaggle.com/code/souravmighty/understanding-audio-data-mel-spectrogram#Bengali.AI-Speech-Recognition:-From-Audio-Signals-to-Mel-Spectrogram
\nHope this will be helpful to get started with this competition.
Thanks. I have just started going through some tutorials on audio signal processing. Best of luck
", "rawMarkdown": "Thanks. I have just started going through some tutorials on audio signal processing. Best of luck", "votes": null }, { "id": "2397092", "postDate": "08/18/2023 17:49:32", "content": "Thanks for the motivation and the resources . Guys like you ! are raising the bars of the community and helping out people here . Thumbs Up !
", "rawMarkdown": "Thanks for the motivation and the resources . Guys like you ! are raising the bars of the community and helping out people here . Thumbs Up !", "votes": null }, { "id": "2403251", "postDate": "08/22/2023 15:20:57", "content": "Really good for ASR starters for 3 reasons:
\n1) The data is small, so training won't cost much time.
\n2) Novices can learn and practice all kinds of ASR models.
\n3) Bengali is just an example, we can apply it to other languages.
this is not for beginners, but I would highly suggest you participate and try to learn going through the discussions and code, if you have the time and interest, its just a different style of competition. best wishes!
", "votes": null, "replies": [ { "id": 2386716, "author_name": "faysalmiah1721758", "author_url": "", "post_date": "08/12/2023 05:56:26", "content": "@anivana I'm a beginner at this stage. I learn and apply Fourier transform and spectrogram. Any suggestion for model building? Which algorithm should I use?
", "votes": null, "replies": [ { "id": 2386847, "author_name": "anivana", "author_url": "", "post_date": "08/12/2023 07:29:47", "content": "https://www.kaggle.com/competitions/bengaliai-speech/discussion/425250
\nStart with this, and you will learn what to do in short time :) best wishes.
", "votes": null, "replies": [] } ] } ] }, { "id": 2385922, "author_name": "supersecurehuman", "author_url": "", "post_date": "08/11/2023 16:20:33", "content": "Hey!
\nThis competition lasts for 2 more months. You can set this as your challenge, and try working towards it. You can start off with available code and previous competition, look for existing models, and various methods used. Assuming you are consistent enough, you should learn a lot after this competition. You will pickup a lot of skills in using Large Models, Training with constraints, Basics of Language Models, How is deep learning used in speech and so forth. Remember to learn about the models you use, not just blindly use them.
\nHappy kaggling!
", "votes": null, "replies": [] }, { "id": 2386707, "author_name": "faysalmiah1721758", "author_url": "", "post_date": "08/12/2023 05:51:34", "content": "@shoabahamed I'm also a beginner, I joined this competition for learning purposes. I suggest you to join this competetion.
", "votes": null, "replies": [] }, { "id": 2387492, "author_name": "msthil", "author_url": "", "post_date": "08/12/2023 17:40:19", "content": "As someone who perpetually participates and never wins but usually learns a lot, 100% yes. Worst case is you've clicked on a button to participate and get free access to data and a GPU on kaggle. The discussions will provide state of the art resources and usually links to courses.
\nI can't remember what competition it was, but I thought I saw a winner one time say they didn't even know what deep learning was when they started lol.
\nNever doubt yourself and as long as it's not too much of a time or mental struggle to affect your life, give it all you got and see where you end up.
", "votes": null, "replies": [] }, { "id": 2390111, "author_name": "snnclsr", "author_url": "", "post_date": "08/14/2023 11:33:33", "content": "I think what matters the most is your motivation. Is this competition excites you in terms of new skills and new ways of approaching problems? If yes, start right away. If you can keep your motivation until the last minute, you'll enjoy a lot as you proceed. And any struggles along the way won't feel like one but a new opportunity to move one step further. So. give it a try, explore the dataset, and decide.
\nIf you want to participate at the end of the day, here are two great resources to start tackling this problem in addition to the posts from the competition host(s) themselves:
\n\n\nhttps://www.youtube.com/playlist?list=PL-wATfeyAMNqIee7cH3q1bh4QJFAaeNv0
\n
\n \n\n
Especially the chapter 5 will give you all the necessary pieces to start with Automatic Speech Recognition
", "votes": null, "replies": [ { "id": 2390318, "author_name": "shoabahamed", "author_url": "", "post_date": "08/14/2023 13:43:55", "content": "Thanks for the suggestions and links.
", "votes": null, "replies": [] }, { "id": 2390368, "author_name": "anivana", "author_url": "", "post_date": "08/14/2023 13:59:10", "content": "great answer and links, thanks!
", "votes": null, "replies": [] }, { "id": 2397092, "author_name": "aibuzz", "author_url": "", "post_date": "08/18/2023 17:49:32", "content": "Thanks for the motivation and the resources . Guys like you ! are raising the bars of the community and helping out people here . Thumbs Up !
", "votes": null, "replies": [] } ] }, { "id": 2391941, "author_name": "souravmighty", "author_url": "", "post_date": "08/15/2023 11:38:08", "content": "I am also a beginner when it comes to nontabular data and hoping to learn something new from this competition. This is the first time I'm working with any audio data and I have been trying to understand the basics of audio signals and how to process it since couple of days.
\nI have summarized all my learning in the following notebook: https://www.kaggle.com/code/souravmighty/understanding-audio-data-mel-spectrogram#Bengali.AI-Speech-Recognition:-From-Audio-Signals-to-Mel-Spectrogram
\nHope this will be helpful to get started with this competition.
Thanks. I have just started going through some tutorials on audio signal processing. Best of luck
", "votes": null, "replies": [] } ] }, { "id": 2403251, "author_name": "berserker408", "author_url": "", "post_date": "08/22/2023 15:20:57", "content": "Really good for ASR starters for 3 reasons:
\n1) The data is small, so training won't cost much time.
\n2) Novices can learn and practice all kinds of ASR models.
\n3) Bengali is just an example, we can apply it to other languages.