Please ask any questions you have about snake classification, taxonomy, biology, ecology, anatomy, or about venomous snakebite here!
There are many corrupt images in the train directory. Is there a way to weed them out?
Nevermind. I found a way to fix them 
There are 181 corrupt files.
Listing them here for others to make use of it.
Images class-67/696c9597141c0820c770dae3f3536eea.jpg class-273/61cf7bb368cfe50e18c86423a2c19e35.jpg class-707/fb0617d72b5f5091b325c294aa4ff7e0.jpg class-561/e470b27eb7e0a141b6777bd475a69496.jpg class-4/200cec355210175fd35b8a313cda5693.jpg class-4/ae72a2493e6537981bef8145aa7a5506.jpg class-4/99a1d0072149ad4657b0224bc4ee40e8.jpg class-4/951f55af6c9481013663d15544b0cfc8.jpg class-4/ce2e90448e4cca480eb8ecf2ed25e116.jpg class-771/428c5b7f70412525dfd0acb47452a53a.jpg class-771/3117568e906087cbede8654118065db3.jpg class-771/2f7e9980f236ee256aa1a679170f3e89.jpg class-771/f7a80eac2f197c34f9dfb5ac8b16d443.jpg class-771/509064a487d9b5ea9e162f6ab5a85bc5.jpg class-771/cb7b4d1646efba3b530d97572664326a.jpg class-771/612379b5f7b7e8f76c72a629649fd6c3.jpg class-771/74e6d64e7a55e3112cb47a45c77a8d9f.jpg class-771/23c72897cc6d1f973d7fb4e0a13bcb5e.jpg class-771/732f99fd84168a20afa3b3dc38a02a04.jpg class-771/f9f7f1f7d276507adf588942be436272.jpg class-771/ddd5880cc925ebdb4e45311da12f681c.jpg class-771/48a81c7c593b4e8866935480417fa63e.jpg class-771/f1f7e59ec8dce6969a022b2da764eaa9.jpg class-771/7e88cefd3a5259ef0b8478e888441aa4.jpg class-771/6f0267472264d42726cd1b4a9e788117.jpg class-771/dca054f6b7ac65c89639328d20c82fe1.jpg class-771/62a356c8e134245f539d5a948736851f.jpg class-771/16c6f99411b78d7675574b7d365891c4.jpg class-771/bfdedbfb857d485141d46b5a7e9046f0.jpg class-771/03fc07ce75eec120a3ca7eeea1e49e08.jpg class-771/cb8228f546c58427e661e92f21a9c6a5.jpg class-771/363ab4b2288f6018270cbbcd1bf801c9.jpg class-390/55418663bf3cfbd0500b944f302ce634.jpg class-390/d421e27a95d555cf418b48c3bd956298.jpg class-390/b8ccbb7e6f42c8f816f0fe0b78472970.jpg class-811/fd3482ce0ec951de46cf0936c6888101.jpg class-811/3e4899b031e353b688f12312b67aecfa.jpg class-72/8b5f0e85cecf5dbba00bca50b0c015e2.jpg class-72/38721cf9a452296dfbe92785725e4576.jpg class-72/ff533de55f1e0c9410fe3fe918b9059c.jpg class-450/8d136064f96d8d256f80880dc8a492ed.jpg class-450/df69240e50ad7d6184f9dcd1ab7a2206.jpg class-450/d2c45b6f1d9c026ae6ba05689d5937c6.jpg class-450/41f31fdcc28d31e423a0c8361cf009bf.jpg class-394/e671dc779eaefb686a5bb8fa7309d2be.jpg class-394/b7c93f4ef3342a711ee731b072f13703.jpg class-536/d19995521d8a9541bbfa344472c3e3d7.jpg class-536/117e6834dc04255884e3a6325543e305.jpg class-536/d175c6b710d00aab32b51a7a422776b8.jpg class-536/6b1776c27ef969094aa0bdcd9c2ee592.jpg class-536/948968d4f4264acedc0ae5f7cdfcb793.jpg class-536/14051b8aa30f5d83ddd95f8bcbf460cd.jpg class-362/509568f31151cc56fc1f7d19ca60c32c.jpg class-362/65a317c22fcb461e3c22ae16c0e8e785.jpg class-362/356c2832ca4a81971101cbe4dca951f1.jpg class-362/ef3971d9f8d7eaeb702e66c89898ea7f.jpg class-362/e76ef00a5a75086ac5360b284332e126.jpg class-362/b8b0a25a4abffbef45ad8cf827efbd4b.jpg class-508/14a16ba4dd693f79d46d53ecede8a747.jpg class-508/fdf6714bd2f88ed18e292c6275bac37f.jpg class-508/ebb55a0d3b2005e6b8d8cc35065c3b0b.jpg class-508/26328f0fd05778077064e79979973933.jpg class-508/031f43d1cc80a10d8889d9edb691b419.jpg class-508/870d01fa18ba018413f945379bf0bcd0.jpg class-508/adbae69549af0d0d1a1afe1f6b9283fc.jpg class-508/0269a06338db632de30ab5dfad9ce275.jpg class-508/47d8d500c3b862c2eed473f89a5aedec.jpg class-508/e5b886cb435c0146f8ecd85c06610b2a.jpg class-1059/3066e251453a0e9dce4acaa6f8441411.jpg class-872/d44b7a7aa395b4ed6db89c40e69fadc6.jpg class-872/45ddf041445a1ea20802f9b0482f2df3.jpg class-872/154fbf0496cc026984fccc3037c43885.jpg class-872/2f7ee5d57fb5c71b9297054050b718ef.jpg class-872/ad71a2d5cb8ab632cc49970d2ac31b3b.jpg class-872/b622898400e438077aa481a8466b7100.jpg class-872/ddbed0515723605f3fde1c3c31f3a5b5.jpg class-872/c65025dafa53daf636c262b9a80076b9.jpg class-663/b7ad76b8ea42c22252abc50020fdaf2d.jpg class-663/e7882c970ed95c860cef5908de64bcd3.jpg class-663/b2fdaab2c721331aa98fec9dc7e7abcf.jpg class-663/73825f2263e30e513c776a8a62ff6d30.jpg class-663/914c05e708718cbec074ffd3ad77ec3a.jpg class-663/bc6e76d443b89590e8accbabca55f474.jpg class-957/6258cfee76e4bfb89b779f0ed6560616.jpg class-957/d416d29da8eba03e750046a41ff86582.jpg class-957/782f2fad8929867ebbaf79d30929aa84.jpg class-957/b92a7d326153fbb41b53a2bbc1d10fdb.jpg class-957/eaae15d2fdbb7544655d79772a24afed.jpg class-654/ce5fea9ac2a8abae6e41be28dad535eb.jpg class-654/32b4a6d571fc9f305675046f07f2220b.jpg class-966/d85a7a20140f224bd8829d506e4a1f3a.jpg class-966/8bf73d52a0f8c9ab1691123c1a09422e.jpg class-966/f2c193488a1dcc85e5cdf308b8b06641.jpg class-239/4f6ffa6908ff2e7f616867f6881683cf.jpg class-597/a41ff022350ab895caf501df9151024d.jpg class-597/505b61540d32f72de4a3f6d9b742c34d.jpg class-1625/28f476f8d5b12a5290a30d214f1369c3.jpg class-326/fa98dd8d9af1b23ff3e63302457f1dc7.jpg class-326/bd0e6341ea35020699ebe2936dc434be.jpg class-78/9ca023d3111a80618259875151d6482b.jpg class-78/6c970024d28eeeef4cbc97a34d45fcc0.jpg class-78/a11458b550b3996700bb4d34e2d0fcc9.jpg class-78/964b2bd444ef083c2526a0645a0fc6f8.jpg class-78/26c28f70ad2200aebab8422db6c7dc51.jpg class-78/7d59966da597a77813be7bee23e43d09.jpg class-78/ae1ad30356998b53abd90376c91bf4d5.jpg class-78/decec33fef0a187b52bbfe4b6ea14906.jpg class-78/000348dd188d957d408420702775dcc4.jpg class-78/cc95a720dcd2e0a9ac8853cb49fa2c22.jpg class-78/81045921cc7a91cce850cc28b6bfd186.jpg class-78/81ca3cbe226578649165a94cbec278c6.jpg class-78/679905e15c1bfdb4ba2291bbeaf0d2b7.jpg class-337/469214bfbc70ceaa606117c357768b9f.jpg class-128/5a6b7ce50bb3c87a52df325fc8143b84.jpg class-639/67d15e737466a88aedf3b4e0c35c6555.jpg class-639/6b7f50400eff0a25f6813edcffc9e0bb.jpg class-639/88d866721b5151adeed06d91e8fcfe27.jpg class-639/fc6185013477961b82ecb728bf918700.jpg class-697/bbc1844d995f4b05a0df69c643fee8af.jpg class-697/5bbba43cfa55650649a08544cd1eecbe.jpg class-697/59689a9e4671e169ce079f8b185b56f4.jpg class-697/fd7769ee18e3461616aef6ddc0b249ba.jpg class-697/eafeeb084f6c6a7261e8fd8a903728e9.jpg class-697/49ec582cf44c5f7f4a20f4d486c28228.jpg class-697/14051479c062bf6e566a65403f958c66.jpg class-697/c4304cf77d9e1b45d9f17fd23fde250b.jpg class-697/ce7ad07733b800ee9c10f06a69b102f8.jpg class-697/b8f53c38efd14768a626612927360248.jpg class-697/c356d9807d36d525a189c01da36b480e.jpg class-804/9a93f8b40d82bed201901629a7c37c3e.jpg class-804/488ec8b64a59f75b6452cb6c3812547c.jpg class-804/df63caf72348a9fec24837246ebaa3de.jpg class-804/435f53eb0dde7039dcc792d08f17fca0.jpg class-804/a7d198cfc6406ac70a16eb09dd3cef2b.jpg class-804/3bdf8dfe7292ec58aa8baa343c4d1e02.jpg class-273/720b4b27d5184d0df7e1c85d5d3bfbae.jpg class-273/5ac108ede9c17d0654f724d256d370c3.jpg class-460/b54c2fad6b7ee67a901df1150e7aa73b.jpg class-460/985b2c10576eedcaf43d80f235b477ee.jpg class-581/46261c46ac476450385383604ef468c1.jpg class-581/67a57971eb8433e82e1809b284d53408.jpg class-581/4dc8ea23f4a3e2a6827696ccdb49edb3.jpg class-140/689cbd4c339155d455025613942d5ff7.jpg class-140/ee396c1f181e7c1d3c9e87279d5912ee.jpg class-140/5e0d366a656cbe7b1ce85b049ba3a51e.jpg class-140/121128312e04b2cd926ab571dde9ede9.jpg class-204/f1850dcb4515872694d18bdbab85d54f.jpg class-204/fd8db1b6101a3b1782ccf4c50be3b4c9.jpg class-204/52a81c5dadf7e2dafb57d1ee394b944d.jpg class-204/cbfe7df6057bb4e2522e418f79d61770.jpg class-204/0b0392d8f7c4f511910fedd3ec125110.jpg class-204/e8a0c7ea3ee1db91c4c032ac228eaaa7.jpg class-204/b4c53457da37a1a85be1f09bf9eb59e2.jpg class-204/30d55f451e450c33fb4da699ec0b6a18.jpg class-204/716c461525068921e9601b882d7119c0.jpg class-204/6974614f3a8bc01b7d8fc9cad18323a0.jpg class-204/9b32534088d83d5b90cf057d167ac4a2.jpg class-204/7d906067a95c8a06c57e1a0c6ca4e0b6.jpg class-204/5326edda0810eed8bcf33ce0560973d7.jpg class-204/e886b740becb075f6aa3e4b9d26dd2b2.jpg class-204/596e17f091e8830e6b55e0ecdc8d2466.jpg class-204/3c2cb0e08fb033259cebd7371e7dc25a.jpg class-204/26aea2334008db1c7f670c9a1dc102d3.jpg class-857/b040c2b1c45274d86804a5a76d34be02.jpg class-857/72cec46d2c2bd835e35bcac0b787d4e4.jpg class-857/fd4aa8ba31536c7dda4398d0b98b025e.jpg class-857/30c0bc03f386fbe049dcd529fb2c34d6.jpg class-857/04de0b194cc002d8d8131abd1ab25a3b.jpg class-857/877475b80e9e5ac4490f4232e95967f3.jpg class-857/70ba8d011d8a2e2bc52d078f14b9edfc.jpg class-543/91931df10498ef05d1d3608e39efba9f.jpg class-543/c0c6b51707b69cd176f50f5068986d0b.jpg class-543/0ab4cd8b3c9061e92960ec0510ffcdc4.jpg class-543/bd6b8bdd14f8dbb5d08b7fda9e9b50a2.jpg class-543/b10e111d78cc30e3fe1e67c0636dc17c.jpg class-526/cf480a068d2e9fbc2bafedc6a6c9a1b7.jpg class-653/7015974c720c7833e098af166523de21.jpg class-635/38b14f0126338683a63c9a1a90e27dec.jpg class-854/498dfa2a80846da28b7149b8a67d1ccc.jpg class-854/3912e63fb92534b40f260ee6b8280bad.jpg class-540/ec1e162b53c89078a273d08938774fdd.jpg
1 Like
Hello,
I am unable to find which class-label corresponds to which snake species. This is a problem for submission as it contains only names of species.
For example, which species do images of class-204 belong to?
Please check image attached and make the list of class-labels and corresponding species available.
Also, there are a number of test images that are corrupt. I have listed them below. Please suggest what needs to be done on these files.
I believe I have a really good model to help this cause. I would like two request you to fastrack these issues.
round1/145e1fed44c0608b2510495e1da186e6.jpg
round1/a781d33863fa2eb233b366982fc311f2.jpg
round1/035b83d6eb12ca399d5fb7a5a4f5b73b.jpg
round1/094684c7f05726dcf48edd40f521de12.jpg
round1/f036432c65478a29ffc288c50c2c862f.jpg
round1/64bbe30099ab1a108db780eeaaf09055.jpg
round1/d98b33a4a9743d41462c423b7a77cf6e.jpg
round1/e808d7e250c6cba745e3ddf6cc7166c4.jpg
round1/06a3288602939c7f03827215e80e1a1e.jpg
round1/d38a244e9f88fbe5d24fab35813e0199.jpg
round1/b099b295efe26cec8064600a6ca5cd97.jpg
round1/8c1caca747f9c9ff4fa85d5deb6a2d7b.jpg
round1/77850402b8b8121d998b8a3b98ed8c57.jpg
round1/2b8060650fbc5c4d04c831e37b6a2af3.jpg
round1/88277e9b21c88548fb99eb32d46b440a.jpg
round1/a7775e93e11bcb9de449b40742fb922f.jpg
round1/bd8e2d98c306b0118abe5294bde8a063.jpg
round1/98726ddfca6ae64e305a31af70d91483.jpg
round1/122231ec40704f3077b5e08d9545e682.jpg
round1/d52ebe0861059d04c0ea16acbbf47648.jpg
round1/f5fe0394d7732da1ce5373133e275874.jpg
round1/142f43fd0a314ecd7f3e4cbfc1f3080c.jpg
round1/9f1acb51f2bbce9e458b1f7593039b14.jpg
round1/1609b108e31ccb318ee8e475ca72074e.jpg
round1/8d0f77893c900842d3f537020601cc81.jpg
round1/b59cbae57050c1b65130c79971309caf.jpg
round1/9998b45c0e5f9db01dcf7fe3d4582d70.jpg
round1/55f787c655671d97b675b3b4a1092e68.jpg
round1/8c57224e5d5d4d76b34f4c467e6c6c73.jpg
round1/23f3c8f701c8fa48f1fce47039a84bb3.jpg
round1/80bf3fd3759e0bbc0b83baf4748284de.jpg
round1/773a7c1b0904561126a608f5fac2a3f8.jpg
round1/79523652fdcea02763665d685b8a96a2.jpg
round1/e21b308876dd40450fd79befadf72e84.jpg
round1/f8675f96b7b71c2c615fc31ca79fd98d.jpg
round1/1c56622f067257a814254b3a0d66a504.jpg
round1/adabb60ac05b85be74ee23c1ed84164b.jpg
round1/3f014dd954ba56697215fcae131f7eac.jpg
round1/2e23ade63c4e32b728a423ff19e52ef1.jpg
round1/e2604125cc5fce0380b826a3c9ad4207.jpg
round1/d8d7b30c4ff40ec049219998f776de19.jpg
round1/7ae396c98d9cf5a4927f89a48be1b027.jpg
round1/5d14c48dc467098b1433dc5ba0d670e7.jpg
round1/2bcb7604d1e6611d3911efb09d0760a6.jpg
@santhosh_shetty: We had the class idx mapping added to the datasets page : https://www.aicrowd.com/challenges/snake-species-identification-challenge/dataset_files
And also, please ignore (add a random prediction) all the files which are corrupted. In the next version of the evaluator. We can have a re-look at the test set data, and remove the corrupt images when computing the final score.
1 Like
Hello,
I have incorporated all the changes as specified. However, I am getting an error saying sum of columns don’t add to 1. I think this is bound to rise when the values are recurring. I have my submission file ready but I am unable to upload due to this issue. Please let me know what could be done.
HI @santhosh_shetty,
The columns represent a probability distribution across all the classes. Hence the sum of the probabilities should be less than 1. On the evaluator, we actually check for the values being less than 1.01 (to leave room for some floating point errors adding up).
Usually people would pass the list of confidences through a softmax function to ensure that the sum of the values is less than 1.
Does this make sense ?
Yes, I did all this but for some reason it does not seem to accept the result and throws out an error saying Please make sure that the sum of each row of predictions is less than 1.
The floating point errors seem to still exist I believe. Please check it. Maybe there is a necessity to slightly increase the range of providing room for floating point errors (All my sums are of the form 0.9999997394239415 which is as close to 1 as it can get). The submission file is ready.
PS : Even the sample submission file fails.
Hey,
I relaxed the prob sub criteria a little bit.
Can you please try submitting again ?
The sample submission does pass now 
1 Like
Hey! Thanks a lot! It passes! We have a valid score now 
Hey, Is it possible to have teams for this challenge?
@santosh_shetty : The site wide teams feature isnt launched yet, and will be launched soon.
In the meantime, the official recommendation is to create a separate account for the team, and only submit from that account.
1 Like
It is possible to know the name of the snake based on its bite on human skin?
1 Like